from legal transplants to legal translations: the ... · americanization thesis in criminal...

Volume 45, Number 1, Winter 2004

From Legal Transplants toLegal Translations:

The Globalization of Plea Bargaining and theAmericanization Thesis in

Criminal Procedure

Máximo Langer∗

I. Introduction

Since the end of the Second World War, and particularly following theend of the Cold War, the American legal system arguably has become themost inºuential legal system in the world.1 American inºuences on the legalsystems of other nations have ranged from general inºuences on jurispruden-tial approaches to law (e.g., legal realism and pragmatism, law and econom-ics, rights discourse, etc.)2 to inºuences on speciªc legal areas (e.g., constitu-

∗ Acting Professor of Law, UCLA School of Law. S.J.D. Candidate, Harvard Law School, 2004; LL.B.,

University of Buenos Aires Law School, 1995.I would like to thank Gianmaria Ajani, Bill Alford, Arthur Applbaum, Tal Ben-Shachar, Scott Cum-

mings, Mirjan Damaška, Sharon Dolovich, Vic Fleisher, Stephen Gardbaum, Latonia Haney, Alon Harel,Ben Hett, Duncan Kennedy, Gia Lee, Alejandro Loritee, Stefano Maffei, Bill McGovern, Martin O’Neill,Sara Olack, Eric Orts, Randy Peeremboom, Kal Raustiala, Arthur Rosett, Gary Rowe, Yoav Sapir, HaniSayed, Roy Schondorf, Patrick Shin, David Sklansky, Carol Steiker, Richard Steinberg, Bill Stuntz, LloydWeinreb, Adam Winkler, and participants in the UCLA School of Law Faculty Workshop, in theSJD/JSD/PhD 2002 Conference at Harvard Law School, and in the conference “Law and Economic De-velopment: Critiques and Beyond” organized by the European Law Research Center at Harvard LawSchool, for helpful comments on earlier drafts. I would especially like to thank Carol Steiker, MirjanDamaška, Phil Heymann, Duncan Kennedy, David Sklansky, and the UCLA School of Law Faculty fortheir encouragement and support. Guillermo Bleichman, Janet Halley, Esther Liberman, and DorisSommer gave me useful bibliographical advice on translation studies. Joanne Jackson and SatyanandSatyanarayana did excellent editing work. Jason Kleiman provided very valuable research assistance. Ialso gratefully acknowledge the support of the Byse Fellowship at Harvard Law School, the HarvardCenter for Ethics and the Professions, the Organization of American States, and the Rockefeller Centerfor Latin American Studies at Harvard University. Myriam Hess, Ben Hett, and Laura Elbert helped mewith some of the German texts. Of course, any mistakes are my responsibility.

1. See, e.g., Duncan Kennedy, Three Legal Globalizations (2001) (unpublished manuscript on ªle withthe Harvard International Law Journal); Ugo Mattei, A Theory of Imperial Law: A Study on U.S. Hegemonyand the Latin Resistance, 10 Ind. J. Global Legal Stud. 383 (2003); Ugo Mattei, Why the Wind Changed:Intellectual Leadership in Western Law, 42 Am. J. Comp. L. 195 (1994); Wolfgang Wiegand, Americanizationof Law: Reception or Convergence?, in Legal Culture and the Legal Profession 137 (Lawrence M.Friedman & Harry N. Scheiber eds., 1996); Wolfgang Wiegand, The Reception of American Law in Europe,39 Am. J. Comp. L. 229 (1991).

2. See, e.g., Kennedy, supra note 1; Mattei, Why the Wind Changed, supra note 1.

2 Harvard International Law Journal / Vol. 45

tional law,3 tax law,4 securities law,5 corporate law,6 patent law,7 interna-tional commercial arbitration,8 etc.); from legal education (e.g., a creditssystem for particular courses, or certain post-graduate studies leading to anLL.M. degree)9 and the structure of the legal profession (e.g., large lawªrms10 or the valorization of private practice11) to the reform of the judici-ary;12 from speciªc legal doctrines or legal tools (e.g., constitutional exclu-sionary rules,13 the doctrine of “actual malice” in the freedom of speech andof the press,14 class actions,15 etc.) to institutional arrangements such as theseparation of powers and judicial review.16 These undeniable American inºu-ences on other legal systems have led a number of commentators, both in theUnited States and abroad, to announce that a substantial number of legalsystems, both at the national and the international levels, may graduallycome to resemble or mimic the American legal system and thus become

3. See, e.g., Constitutionalism and Rights: The Inºuence of the United States Constitu-

tion Abroad (Louis Henkin & Albert J. Rosenthal, 1990); Joan Davison, America’s Impact on Constitu-tional Change in Eastern Europe, 55 Alb. L. Rev. 793 (1992). For a work cautioning against the adoptionof the U.S. constitutional model by other countries, see Jonathan Zasloff, The Tyranny of Madison, 44UCLA L. Rev. 795 (1997).

4. Miranda Stewart, Global Trajectories of Tax Reform: The Discourse of Tax Reform in Developing andTransition Countries, 44 Harv. Int’l L.J. 139, 151 (2003).

5. See, e.g., Frederick Tung, Passports, Private Choice, and Private Interests: Regulatory Competition and Co-operation in Corporate, Securities, and Bankruptcy Law, 3 Chi. J. Int’l L. 369 n.43 (2002).

6. Paul von Nessen, The Americanization of Australian Corporate Law, 26 Syracuse J. Int’l L. & Com.

239 (1999).

7. See, e.g., Julia Alpert Gladstone, Why Patenting Information Technology and Business Methods is NotSound Policy: Lessons from History and Prophecies for the Future, 25 Hamline L. Rev. 217, 229 (2002).

8. Yves Dezalay & Bryant G. Garth, Dealing in Virtue: International Commercial Arbi-

tration and the Construction of a Transnational Legal Order (1996).9. See, e.g., Wiegand, Americanization, supra note 1, at 138–39; Wiegand, The Reception, supra note 1, at

232–33.10. See, e.g., Note, The Protectionist Bar against Foreign Lawyers in Japan, China and Korea: Domestic Con-

trol in the Face of Internationalization, 16 Colum. J. Asian L. 385, 389 (2003) (“The growing presence oflaw ªrms in Japan has encouraged a further Americanization of legal practice, but not without continu-ing restrictions.”).

11. See, e.g., Bernard Michael Ortwein II, The Swedish Legal System: An Introduction, 13 Ind. Int’l &

Comp. L. Rev. 405 (2003) (n.247: “Professor Modeer maintains that ‘Americanization’ of the Swedishlegal profession has resulted in an increase in the number of law school graduates who seek membershipinto the Swedish Bar Association, hoping to facilitate a career in the private sector of law practice.”).

12. See, e.g., Hiram E. Chodosh, Reforming Judicial Reform Inspired by U.S. Models, 52 DePaul L. Rev.

351 (2002).13. Argentina, Canada, Italy, and Spain, among other countries, have adopted a version of the Ameri-

can exclusionary rule regarding evidence obtained in violation of constitutional rights. On the Canadianexperience, see, for example, James Stribopoulos, Lessons from the Pupil: A Canadian Solution to the AmericanExclusionary Rule Debate, 22 B.C. Int’l & Comp. L. Rev. 77 (1999).

14. For an analysis of the reception of this doctrine by the Supreme Court of Argentina, see, for exam-ple, Eduardo A. Bertoni, Libertad de expresión en el estado de derecho 72–82 (2000).

15. Antonio Gidi, Class Actions in Brazil—A Model for Civil Law Countries, 51 Am. J. Comp. L. (2003)

(pointing out how Anglo-American class actions inºuenced Brazilian developments through the work ofItalian scholars).

16. See, e.g., Erhard Blankenburg, Changes in Political Regimes and Continuity of the Rule of Law in Ger-many, in Herbert Jacob et al., Courts, Law, and Politics in Comparative Perspective 249, 308

(1996) [hereinafter Courts, Law, and Politics].

2004 / Legal Translations of American Plea Bargaining 3

“Americanized.”17 Other commentators, while acknowledging the predomi-nant inºuence of the American legal system, have stopped short of assertingthat American inºuence is actually recreating American legal practice innon-American jurisdictions.18

In this Article, I caution against the former thesis of Americanization (the“strong” thesis) through an examination of the introduction of American-style plea bargaining in four civil law countries—Germany, Italy, Argentina,and France.19 The inºuence of American plea bargaining in all four of thesejurisdictions is undeniable. Despite this inºuence, however, the importationof plea bargaining into these jurisdictions is not likely to reproduce an Ameri-can model of criminal procedure. Each of these jurisdictions has adopted aform of plea bargaining that contains differences—even substantial differ-ences—from the American model, either because of decisions by the legalreformers in each jurisdiction or because of structural differences betweenAmerican criminal procedure and the criminal procedures of the civil lawtradition. Consequently, some civil law versions of plea bargaining have notresembled the American practice since their inception. In addition, thestructural differences between the American adversarial conception of crimi-nal procedure and the continental European and Latin American inquisito-rial conception of criminal procedure are so deep that individual reformsinspired by American models are unlikely to push these inquisitorial crimi-nal procedures in the direction of the American adversarial system. Finally,in each of these civil law jurisdictions, some legal actors have distrusted orresisted the adoption of plea bargaining and other consensual mechanisms,either because reforms have threatened their traditional powers within theinquisitorial criminal process or because of their differing legal culture. Thisdistrust and resistance has also played a role in neutralizing the potential forAmericanization inherent in some of these criminal procedure reforms.

Although the factors hindering Americanization have been present ineach of the civil law jurisdictions examined, they have played out differentlyin different jurisdictions. Not only has each of these jurisdictions adopted aversion of plea bargaining different from the American model, but also, each

17. See, e.g., R. Daniel Keleman & Eric C. Sibbitt, The Americanization of Japanese Law, 23 U. Pa. J.

Int’l Econ. L. 269 (2002); Paul von Nessen, supra note 6; Wiegand, The Reception, supra note 1, at 246–48 (comparing the “Americanization” of current European legal systems to the reception of the ius com-mune in the Middle Ages in that continent).

18. See, e.g., Gidi, supra note 15; Kennedy, supra note 1; Stephen Zamora, The Americanization of Mexi-can Law: Non-Trade Issues in the North American Free Trade Agreement, 24 Law & Pol’y Int’l Bus. 391

(1993) (discussing how Mexico’s labor and environmental laws, and related enforcement mechanisms,differ from those of the United States, despite pressures on the Mexican legal system to mimic that of theUnited States).

19. I have chosen these four jurisdictions for two reasons. First, they have traditionally been consid-ered inºuential on other countries (France, Germany, and Italy worldwide, and Argentina in LatinAmerica). Thus, to the extent that they still retain part of their ascendance over other civil law jurisdic-tions, they may provide a good proxy as to what may happen in a good part of the civil law. Second, eachof these jurisdictions provide a different version of how plea bargaining may be imported by civil lawcountries and what different effects these different versions may produce in each of them.


one of these jurisdictions has adopted forms of plea bargaining different fromone another. The German, Italian, Argentine, and French plea bargains dif-fer substantially amongst themselves because of decisions by legal reformersin each of these countries, the differing ways in which the practice has beenintroduced, and the resistance it has generated. Given the differences amongthese plea bargains, the adoption of some form of plea bargaining in thesejurisdictions may produce different transformations or effects in each juris-diction. Therefore, the potential inºuence of American plea bargaining oncivil law jurisdictions may not be that civil law systems will gradually re-semble the American legal system, but rather that they may begin to differamongst themselves in aspects on which, until very recently, they have beenrelatively homogeneous. In other words, the paradoxical effect of Americaninºuence on the criminal procedures of the civil law tradition may not beAmericanization, but rather fragmentation and divergence within the civillaw.

Before beginning the concrete analysis of these importations, it is neces-sary to address two issues. First, in order to explain and trace the transforma-tions that American plea bargaining has undergone since ªrst adopted bythese civil law jurisdictions, it is necessary to explain the historical differ-ences between the American adversarial criminal procedure and the inquisi-torial criminal procedures of the civil law tradition. The adversarial and theinquisitorial systems can be understood not only as two different ways todistribute powers and responsibilities between various legal actors—the de-cision-maker (judge and/or jury), the prosecutor, and the defense—but alsoas two different procedural cultures and thus, two different sets of basic un-derstandings of how criminal cases should be tried and prosecuted.20 Themajor difference between the procedural cultures for the purposes of thisArticle is that, whereas the adversarial system conceives criminal procedureas governing a dispute between two parties (prosecution and defense) beforea passive decision-maker (the judge and/or the jury), the inquisitorial systemconceives criminal procedure as an ofªcial investigation, done by one ormore impartial ofªcials of the state, in order to determine the truth.21

20. In this sense, it is important to emphasize from the outset that I will use the expression “adver-

sarial system” as a descriptive category, not as a normative ideal. As a normative ideal, the expression issometimes used in the United States to refer to a criminal procedure where the rights of the defendant arefully respected, see, e.g., Mirjan Damaška, Adversary System, 1 Encyclopedia of Crime and Justice

24, 25 (Sanford H. Kadish ed., 1983), and the epitome of the adversarial system is the trial by jury.However, in this Article, I will use the expression “adversarial system” as a descriptive category throughwhich I will explain the current features of American criminal procedure in opposition to the currentfeatures of criminal procedure in continental Europe and Latin America. Similarly, the expression “in-quisitorial system” is sometimes used in a negative way to refer to authoritarian conceptions of criminalprocedure. But in this Article, I will use the expression “inquisitorial system” only as a descriptive category.

21. As I explain later, I do not conceive of the adversarial and inquisitorial systems as homogenousprocedural cultures. But these two conceptions have been predominant in the ªve jurisdictions I willexamine in this Article, and thus they are the most relevant for our study.


Given that these cultural differences have been overlooked by traditionalcomparative criminal procedure scholarship, I propose in this Article a newtheoretical framework to reconceptualize the adversarial and the inquisitorialsystems. This new theoretical framework will be useful not only to describethe differences between the criminal procedures of the common and civil lawtraditions but also to analyze some of the transformations that plea bargain-ing—and potentially other legal institutions—may undergo when trans-ferred from one system to the other. This theoretical framework will also beuseful in assessing Americanization, as brought about by the adoption ofsome form of plea bargaining in these four civil law criminal procedural sys-tems, because it will provide a clear axis of reference in comparing the dif-ferences between the adversarial and the inquisitorial systems, and thus, inevaluating whether the latter are moving in the direction of the former.

Second, it is necessary to question and reassess how we think about thecirculation of legal ideas between legal systems. The metaphor of the “legaltransplant” has been the main device used by comparative law scholars andpractitioners when analyzing the importation of foreign legal practices.22

This metaphor, however, has its shortcomings. Its chief problem is that itconveys the notion that legal ideas and institutions can simply be “cut andpasted” between legal systems. Thus, this metaphor fails to account for thetransformation that legal ideas and institutions may undergo when they aretransferred between legal systems. In this Article, I propose the metaphor ofthe “legal translation” as an alternative heuristic device to employ whenanalyzing the transfer of legal ideas and institutions between legal systems.The adversarial and the inquisitorial systems, understood as two differentprocedural cultures, can be understood as two different systems of produc-tion of meaning. Thus, the transfer of legal institutions from one system tothe other can be understood as translations from one system of meaning tothe other. Speciªcally, the transformations that plea bargaining has under-gone when transferred to these civil law jurisdictions can be understood ei-ther as decisions taken by the “translators” (i.e., legal reformers) or as a

22. See generally Daniel Berkowitz et al., Economic Development, Legality, and the Trans-

plant Effect (Center for International Development at Harvard University, Working Paper No. 39,2000, available at http://www.cid.harvard.edu/cidwp/039.htm (last visited Nov. 25, 2003) (on ªle withthe Harvard International Law Journal); Edgardo Buscaglia & William Raºiff, Law and Eco-

nomics in Developing Countries 14–19, 31–54 (2000); Alan Watson, Legal Transplants: An

Approach to Comparative Law 21 (2d ed. 1993); Gianmaria Ajani, By Chance and Prestige: LegalTransplants in Russia and Eastern Europe, 43 Am. J. Comp. L. 93 (1995); Ugo Mattei, Efªciency in LegalTransplants: An Essay in Comparative Law and Economics, 14 Int’l Rev. L. & Econ. 3 (1994); Jonathan M.Miller, A Typology of Legal Transplants: Using Sociology, Legal History and the Argentine Example to Explain theTransplant Process, 51 Am. J. Comp. L. (forthcoming Fall 2003) (on ªle with Harvard International LawJournal); Loukas A. Mistelis, Regulatory Aspects: Globalization, Harmonization, Legal Transplants and LawReform—Some Fundamental Observations, 34 Int’l Law. 1055 (2000); Frederick Schauer, The Politics andIncentives of Legal Transplantation, in Governance in a Globalizing World, 253 (Joseph S. Nye Jr. &John D. Donahue eds., 2000); Jonathan B. Weiner, Responding to the Global Warming Problem: SomethingBorrowed for Something Blue: Legal Transplants and the Evolution of Global Environmental Law, 27 Ecology

L.Q. 1295 (2001).


product of the structural differences that exist between the adversarial andinquisitorial “languages.”

In this Article, I show that the inºuence of American criminal procedureseems to conªrm the weak version of the Americanization thesis and that thestrong version of the Americanization thesis is inapplicable, or, at least,overly simplistic, with respect to criminal procedure. Despite the inºuenceof American plea bargaining on civil law jurisdictions, the cultural differ-ences between the adversarial and the inquisitorial systems are too deep tobe overcome by a single American-inspired reform or even a substantialnumber of American-inspired reforms. Furthermore, given that each of thesecivil law jurisdictions has translated plea bargaining in a different way, theultimate effect of this American inºuence may end up being a fragmentationand divergence, rather than Americanization, within civil law criminal pro-cedure. The four jurisdictions examined in this Article share a similar pre-dominant procedural culture that regards criminal procedure as an ofªcialinvestigation conducted by impartial ofªcials. However, each jurisdictionhas adopted a different version of plea bargaining, and thus, these differentreforms may lead them in different directions.

The structure of the Article is as follows. In Parts II and III, I propose anew theoretical framework to conceptualize the adversarial and the inquisi-torial systems. The main idea that I defend here is that these categoriesshould be understood not only as two different ways to arrange powers andresponsibilities between the main actors of the criminal process (judges,prosecutors, defense attorneys, etc.), but also as two different procedural cul-tures. Part IV shows the inºuences that the Anglo-American conception ofcriminal procedure has had on the inquisitorial criminal procedure of conti-nental Europe and Latin America in recent decades. Part V discusses theproblems presented by the metaphor of the “legal transplant” as a heuristicdevice in analyzing the circulation of legal institutions between legal sys-tems. In its place, I offer the metaphor of the “legal translation” as a morenuanced and productive heuristic device to think about these issues. Thetransferred legal practice—plea bargaining in this instance—can be thoughtof as the “text” that has been translated from one “language”—the adversar-ial system of the United States—to another “language”—the inquisitorialsystems of Germany, Italy, Argentina, and France. Part VI explains how pleabargaining could have a potential Americanization effect on civil law juris-dictions if fully accepted by these systems. My argument here is that Ameri-can plea bargaining assumes an adversarial conception of criminal procedure,and thus, if faithfully translated and fully accepted by civil law jurisdictions,could advance the American conception of criminal procedure in inquisito-rial jurisdictions. Parts VII, VIII, IX, and X analyze the German, Italian,Argentine, and French “plea bargains,” and also explain not only why eachof these countries has translated American plea bargaining but also why eachof them has chosen to translate the practice in a different manner. Here, Ialso demonstrate how the inquisitorial cultural conception of criminal pro-


cedure has been one of the central reasons why American plea bargains havebeen both transformed and resisted in these jurisdictions. Finally, Part XIdraws together my central argument: Given that each of the jurisdictionsexamined has translated plea bargaining in a different way, American inºuencesmay end up producing the fragmentation and divergence, rather than theAmericanization of the criminal law procedures of the civil law tradition.

II. Re-conceptualizing the Adversarial and the

Inquisitorial as Theoretical Categories

Traditionally, scholars of comparative law and comparative criminal pro-cedure have employed the theoretical categories of “adversarial” and “inquisi-torial” through two primary approaches:23 the lowest-common-denominatorapproach and the ideal-type approach.24

According to the lowest-common-denominator approach, the adversarialand the inquisitorial categories simply contain the features common to allcriminal procedure systems of the common and civil law, respectively.25 Forinstance, the trial by jury or the hearsay rule would be features of the adver-sarial system if all common law jurisdictions included these elements at acertain moment in time. This approach presents several problems. First,each time a particular jurisdiction leaves aside a feature that was part of thisminimum common denominator, the proponents of the approach have thefollowing dilemma: either they have to modify the content of the adversarialand the inquisitorial categories, or they have to remove this jurisdictionfrom either the adversarial or the inquisitorial system. The lowest-common-denominator approach does not provide a clear answer for dealing with thisdilemma.26 Second, it is also not clear how to distinguish between the pri-

23. The use of the expressions “adversarial” and “inquisitorial” has not been limited to comparative

criminal procedure. For descriptions of other uses of these terms, see Mirjan Damaška, Adversary System,in 1 Encyclopedia of Crime and Justice 24 (Sanford H. Kadish ed., 1983); Malcolm Feeley, TheAdversary System, in Encyclopedia of the American Judicial System 753 (Robert J. Janosik ed.,1987); Máximo Langer, La Dicotomía acusatorio-inquisitivo y la importación de mecanismos procesales de latradición juridical anglosajona. Algunas reºecciones a partir del procedimiento abreviado, in El Procedimiento

Abreviado 97, 102–11 (Julio B. J. Maier & Alberto Bovino eds., 2001).For studies of other models and categories that have been used for analyzing different types of criminal

procedure, see, for example, Herbert L. Packer, The Limits of the Criminal Sanction 149–246

(1968); Peter Arenella, Rethinking the Functions of Criminal Procedure: The Warren and Burger Courts’ Com-peting Ideologies, 72 Geo. L.J. 183 (1983); Mirjan Damaška, Models of Criminal Procedure, 51 Zbornik

(Collected Papers of Zagreb Law School) 477 (2001); John Grifªths, Ideology in Criminal Procedure,or a Third “Model” of the Criminal Process, 79 Yale L.J. 359 (1970); Erik G. Luna, The Models of CriminalProcedure, 2 Buff. Crim. L. Rev. 389 (1999).

24. See generally Mirjan R. Damaška, The Faces of Justice and State Authority 4–6 (1986).25. See Mirjan Damaška, Adversary System, supra note 20, at 28 (giving as an example of this approach;

Joachim Herrmann, Various Models of Criminal Proceedings, 2 S. Afr. J. Crim. L. & Criminology 3, 4–6(1978). In order to reject the idea that the exclusion from trial of illegally obtained evidence is not adeªning feature of the adversarial approach, Prof. Herrmann mentions that the exclusionary rule has notbeen adopted by all Anglo-American jurisdictions, and that it has been adopted in the legislation ofseveral European countries. Id. at 18.)

26. See Damaška, supra note 24, at 5.


mary and secondary features of both systems. In other words, the lowest-common-denominator approach does not provide an answer to why certainfeatures are more relevant than others in distinguishing the two systems.27

Third, this approach does not seem useful in analyzing hybrid systems,which have been inºuenced by both legal traditions, and perhaps also byother legal traditions of the world. These mixed systems are very importantphenomena to study now, not only because the incidence of hybrid nationalsystems is greater than in the past due to mutual inºuences between thecommon and the civil law, but also in order to understand internationalcriminal tribunals, from Nuremberg to the International Criminal Court,where rules and legal professionals from different legal traditions meet andinteract.28 Finally, this approach does not seem to provide a useful theoreti-cal framework to deal with the phenomena of legal inºuences and legaltransplantation because the most this approach can offer when a legal prac-tice (such as plea bargaining) is transferred from an adversarial to an inquisi-torial jurisdiction is either that the absence of this institution is no longer acharacteristic feature of the inquisitorial system, or that the speciªc receiv-ing jurisdiction is no longer a part of the inquisitorial legal family, thusleading us back to the ªrst problem of the lowest-common-denominatorapproach.

The second main approach to comparative criminal procedure has been toconceptualize the adversarial and the inquisitorial as Weberian ideal-types.29

These models do not exactly exist in any historical legal system, but whilethe common law jurisdictions would be closer to the adversarial type, thecivil law jurisdictions would be closer to the inquisitorial type.30

This theoretical conception is much more promising than the lowest-common-denominator approach and addresses many of its shortcomings. For

27. See id.28. For analyses of the procedure and rules of evidence of the different international criminal jurisdic-

tions based on the dichotomy of the adversarial and the inquisitorial, see, for example, Guillaume Champy,Inquisitoire-Accusatoire devant les juridictions pénales internationales, 68 Int’l Rev. Penal L. 149 (1997);Nancy Amoury Combs, Copping a Plea to Genocide: The Plea Bargaining of International Crimes, 151 U. Pa.

L. Rev. 1 (2002); Richard May & Marieke Wierda, Trends in International Criminal Evidence: Nuremberg,Tokyo, The Hague, and Arusha, 37 Colum. J. Transnat’l. L. 725 (1999); Daryl A. Mundis, From ‘CommonLaw’ Towards ‘Civil Law’: The Evolution of the ICTY Rules of Procedure and Evidence, 14 Leiden J. Int’l L.

367 (2001); Vladimir Tochilovsky, Rules of Procedure for the International Criminal Court: Problems to Addressin Light of the Experience of the Ad Hoc Tribunals, 46 Neth. Int’l L. Rev. 343 (1999).

29. In his classic description of Weber’s methodology, Max Rheinstein writes:Situations of such “pure” type have never existed in history. They are artiªcial constructs similar tothe pure constructs of geometry. No pure triangle, cube, or sphere has ever existed. But never couldreality have been penetrated scientiªcally without the use of the artiªcial concepts of geometry. Forthe “pure” concepts created by him, Weber used the term “ideal type” . . . . The “ideal types” . . .are simply mental constructs meant to serve as categories of thought the use of which will help us tocatch the inªnite manifoldness of reality by comparing its phenomena with those “pure” typeswhich are used, so to speak, to serve as guide in a ªling system.

Max Rheinstein, Introduction to Max Weber on Law in Economy and Society XXIX–XXX (EdwardShils & Max Rheinstein trans., Max Rheinstein ed., 1954).

30. Mirjan Damaška has shown the most sophisticated use of this approach, though he replaces theadversarial and the inquisitorial with his own categories. See Damaška, supra note 24.


instance, the use of ideal-types identiªes the relevant differences between theadversarial and inquisitorial systems, and a change to the concrete criminalprocedure of a system does not necessarily lead to changing or discarding themodel. Instead, this approach only labels concrete criminal procedure ascloser to or farther from the ideal-type. Additionally, the ideal-type approachmay also be fruitful in analyzing hybrid criminal procedures like the inter-national criminal tribunals by identifying those features that correspondwith each type.31 The ideal-type approach also provides a useful frameworkfor analyzing the phenomena of inºuences and legal transplantation betweencommon and civil law procedures.32 First, by providing a clear axis of refer-ence, this approach allows us to analyze the extent to which criminal proce-dures are transformed as a consequence of inºuences or transplantation. Sec-ond, to the extent that the concrete criminal procedures are closer to theideal types, they explain the incentives and logic of the system so that it ispossible to identify in advance potential foci of resistance to reform.

A. Structures of Interpretation and Meaning

Nevertheless, there is one area the ideal-type approach does not address,or leaves in the background, and I would like to bring it to the foregroundin this Article. The adversarial and the inquisitorial models are not only twoways to structure criminal proceedings, but also two ways of understandingand representing criminal procedure. For instance, common law judges par-ticipate in the interrogation of witnesses much less than do their Continen-tal colleagues, not only because procedural rules give them less power to doso, but also because the role of the judge is understood differently in thecommon law system. Whereas the inquisitorial system judge is understoodand perceived as an active investigator with, consequently, the duty to beactive in these interrogations, the adversarial system judge is usually under-stood as a passive umpire who is not supposed to participate actively in theinterrogation of witnesses.33

31. See, e.g., Tochilovsky, supra note 28.32. See Mirjan Damaška, The Uncertain Fate of Evidentiary Transplants: Anglo-American and Continental

Experiences, 45 Am. J. Comp. L. 839 (1997); Elisabetta Grande, Italian Criminal Justice: Borrowing andResistance, 48 Am. J. Comp. L. 227 (2000).

33. With the re-conceptualization of the adversarial and the inquisitorial systems developed in thisArticle, I am trying to capture the internal point of view of the legal actors who are part of criminalprocedures of the common and the civil law. This distinction between an external and internal point ofview has been used in a multiplicity of disciplines. For instance, Max Weber took this distinction intoaccount by including as the object of sociological study not only the external aspect of human behaviorbut also the meaning that human beings ascribe to their own behavior. See Rheinstein, supra note 29, atXXIX–XXX:

The observation that one man hands to another a piece of greenish paper is as such irrelevant in thestudy of human relationships. The observed phenomenon does not assume social signiªcance untilwe know that a large group of human beings, of which our two actors are members, regard thegreenish paper as a piece of paper money or, in other words, that they ascribe to it the function ofserving as a generally recognized means of exchange and payment.

Id. at XX. In jurisprudence, the classic reference is H. L. A. Hart, who used this distinction between


In other words, the adversarial and the inquisitorial can be understood astwo different structures of interpretation and meaning through which the actorsof a given criminal justice system understand both criminal procedure andtheir role within the system.34 Within these two procedural structures ofinterpretation and meaning or “procedural languages,” the same terms orsigniªers often have different meanings. For instance, in the adversarial sys-tem, the word “prosecutor” means a party in a dispute with an interest atstake in the outcome of the procedure; in the inquisitorial system, however,the word signiªes an impartial magistrate of the state whose role is to inves-tigate the truth.35 The word “truth” also has a different meaning in eachprocedural structure of interpretation and meaning. In the adversarial sys-tem, even if the dispute is about “truth,” the prosecution tries to prove thatcertain events occurred and that the defendant participated in them, whilethe defense tries to question or disprove this attempt. The adversarial con-ception of truth is more relative and consensual: if the parties come to anagreement as to the facts of the case, through plea agreements or stipula-tions, it is less important to determine how events actually occurred.36 In theinquisitorial structure of interpretation and meaning, “truth” is conceived inmore absolute terms: the ofªcial of the state—traditionally, the judge—issupposed to determine, through an investigation, what really happened,

external and internal point of view. H. L. A. Hart, The Concept of Law 88–91 (2d ed. 1994). See alsoPaul W. Kahn, The Cultural Study of Law (1999). In anthropological studies, Clifford Geertz hastried to capture the idea of the internal point of view through his concept of “thick description.” Clif-

ford Geertz, The Interpretation of Cultures 5–10 (1973). Since Weber coherently used theideal-type approach and took into account the internal point of view of the social actors, I do not considermy re-conceptualization of the adversarial and inquisitorial systems as incompatible with the ideal-typeapproach; I think they are simply independent from one another. My point is that comparative criminalprocedure analyses have generally overlooked the internal point of view of the actors of the criminaljustice system and that including this perspective is central both to examining the differences betweenthe criminal procedures of the common and the civil law as well as phenomena such as the circulation oflegal ideas and institutions between them and to debates about the Americanization of law.

34. I take the expression “structures of meaning” from Clifford Geertz, Local Knowledge: Fact and Lawin Comparative Perspective, in Local Knowledge: Further Essays in Interpretive Anthropology

167, 182 (1983) (which states: “The turn of anthropology . . . toward heightened concern with structuresof meaning in terms of which individuals and groups of individuals live out their lives, and more par-ticularly with the symbols and systems of symbols through whose agency such structures are formed,communicated, imposed, shared, altered, reproduced, offers as much promise for the comparative analysisof law as it does for myth, ritual, ideology, art, or classiªcation systems, the more tested ªelds of its ap-plication.” Id.)

35. On the different conceptions of the prosecutor in the Anglo-American system and the inquisito-rial one, see, for example, William T. Pizzi, Understanding Prosecutorial Discretion in the United States: TheLimits of Comparative Criminal Procedure as an Instrument of Reform, 54 Ohio St. L.J. 1325, 1349–51(1993); Thomas Weigend, Prosecution: Comparative Aspects, in 3 Encyclopedia of Crime & Justice

1232, 1233–34 (Joshua Dressler ed., 2d ed. 2002).36. This may sound like an exaggeration because, in U.S. jurisdictions, the judge still has to verify the

factual basis for a guilty plea. But in practice, U.S. judges are usually deferential to the agreements of theparties about the facts. Furthermore, cases such as North Carolina v. Alford, 400 U.S. 25 (1970), wherethe defendant pleaded guilty while claiming that he was actually innocent, are clear examples of a con-ception of truth that is much more relative and consensual.


regardless of the agreements or disagreements that prosecution and defensemay have about the event.37

At the same time, there are certain ideas and concepts that exist in oneprocedural language and not in the other. For instance, the adversarial sys-tem includes both the concepts of “confession”—i.e., an admission of guiltbefore the police—and “guilty plea”—an admission of guilt before the courtthat, if accepted, has as its consequence the end of the phase of determina-tion of guilt or innocence. The inquisitorial procedural structure of interpre-tation and meaning or “language,” on the other hand, does not include theconcept of the “guilty plea”; it only includes the concept of “confession.”38

In this system, a defendant cannot end the phase of determination of guilt orinnocence by admitting his guilt before the court. While the admission ofguilt may be very useful to the judge in seeking the truth, the judge still hasthe ªnal word on the determination of guilt and can thus say, “I believe you,but your confession alone is not proof beyond a reasonable doubt that youdid it.” In any case, if an admission of guilt happens during the pre-trialphase, the case must still go to trial before the judge can make a ªnal deter-mination.39

It is important to emphasize that even if it is possible to analyze the ad-versarial and the inquisitorial structures of interpretation and meaning asabstract systems, these structures only exist within concrete criminal proce-dures. In other words, it is possible to identify adversarial and inquisitorialsystems because the legal actors of the Anglo-American and civil law juris-dictions constantly make use of adversarial and inquisitorial structures ofinterpretation and meaning in conscious and unconscious ways. Taking adistinction from structuralist linguistics, I could say that while the struc-tures of interpretation and meaning are the langue—the abstract system ofdifferences that establishes a lexicon and a set of potential operations—thepractices are the parole—the actual speech acts made possible by the langue.40

B. Individual Dispositions

Just as adversarial and inquisitorial structures of interpretation and meaningare grounded in concrete procedural practices, they are also internalized bythe relevant legal actors. I call this the “dimension of individual disposi-

37. On the conception of truth predominant in the inquisitorial system as opposed to the accusatorial

one, see Antoine Garapon, French Legal Culture and the Shock of ‘Globalization,’ 4 Soc. & Legal Stud. 493,

496–97 (1995).38. See, e.g., Myron Moskovitz, PERSPECTIVE: The O.J. Inquisition: A United States Encounter with

Continental Criminal Justice, 28 Vand. J. Transnat’l L. 1121, 1153 (1995). This Article describes thedifferences between the adversarial and the inquisitorial systems by developing a dramatization of how acase similar to the O. J. Simpson trial might be handled by a civil law European criminal justice system.

39. See, e.g., John H. Langbein, Comparative Criminal Procedure: Germany 73–74 (1977).40. See Ferdinand de Saussure, Course in General Linguistics (Wade Baskin trans., Charles

Bally & Albert Reidinger eds., 1959).


tions.”41 These internal dispositions are acquired by the internalization ofthe procedural structures of interpretation and meaning, through a numberof socialization processes (i.e., law schools, judiciary school, prosecutor’sofªce and law ªrm training, interaction with the courts, etc.).42 As a resultof this socialization, a substantial number of actors in the criminal justicesystem are predisposed to understand criminal procedure and the variousroles within it in a particular way, and these dispositions become durableover time.

Hence, the idea that the judge is supposed to be a passive umpire in theadversarial system is not only due to the adversarial structure of interpreta-tion and meaning; it is also due to the phenomenon that a substantial num-ber of legal actors have internalized this structure of meaning in a commonlaw jurisdiction, they have come to consider this as the proper role of ajudge and will usually act accordingly—i.e., censoring a judge who partici-pates too actively in the interrogation of witnesses.43 In other words, to theextent that legal actors internalize these structures of meaning and then in-terpret and interact with reality through them, one could say that thesestructures of meaning constitute and shape legal actors as subjects.44

The dimension of individual dispositions becomes particularly importantwhen studying the transfer of legal ideas, norms, and institutions betweenadversarial and inquisitorial systems, as well as legal transplants in general.As we will see, part of the literature of the legal transplant discusses thetransfer of legal ideas from the common to the civil law, or vice versa, exclu-sively as interactions and struggles between two abstract systems of mean-ing.45 By including this dimension of individual dispositions, I would liketo emphasize that besides these interactions between two abstract systems—that again, as I conceive of them, are inscribed in concrete social practices—every legal transplant also involves interactions between concrete peoplewith a concrete set of individual dispositions.

41. My source of inspiration for the development of this dimension of internal dispositions is sociolo-

gist Pierre Bourdieu’s concept of habitus which can be deªned as “a set of dispositions which induce agentsto act and react in certain ways. The dispositions generate practices, perceptions, and attitudes which are‘regular’ without being consciously co-ordinated or governed by any ‘rule.’” John B. Thompson, Editor’sIntroduction to Pierre Bourdieu, Language and Symbolic Power 1, 12 (Gino Raymond & MatthewAdamson trans., John B. Thompson ed. 1991). Pierre Bourdieu describes his notion of habitus in otherworks as well. Pierre Bourdieu, Raisons Pratiques: Sur La Théorie de l’Action 22–23 (1994);Pierre Bourdieu, Some Properties of Fields, in Sociology in Question 72 (Richard Nice trans., 1993). Isay that the dimension of internal dispositions is only inspired by Bourdieu’s concept, because I do notfollow his theoretical framework in this paper. Thus, my use of it is idiosyncratic.

42. For a description of how lawyers, judges, prosecutors, and professors are trained and socialized in civil lawcountries, see, for example, John Henry Merryman, The Civil Law Tradition 101–10 (2d ed. 1985).

43. On the role of the judge in the United States during trial, see, for example, Craig M. Bradley,United States, in Criminal Procedure: A Worldwide Study 395, 421–22 (Craig M. Bradley ed., 1999).

44. For a recent analysis—in the context of the debate in the United States on “broken windows” po-licing—of how criminal law and criminal justice practices can shape social subjects, see Bernard E.

Harcourt, Illusion of Order 160 (2001).45. See, e.g., Gunter Teubner, Legal Irritants: Good Faith in British Law or How Unifying Law Ends Up in

New Divergences, 61 Mod. L. Rev. 11 (1998).


Moreover, this dimension is also important in understanding how proce-dural systems—and any other structures of interpretation and meaning—can change over time. The relationship between procedural structures ofmeaning and individual dispositions is one of mutual inºuence. Individualdispositions are inºuenced by structures of meaning to the extent that indi-vidual actors acquire, through a number of socialization processes, a concretesystem of interpretation that, to a certain extent, “sets them up” to act andreact in certain ways. But if a signiªcant number of actors internalize a dif-ferent structure of meaning (i.e., when civil law lawyers study in the UnitedStates and partly acquire an adversarial structure of meaning), this may leadthem to question the predominant structure of meaning (i.e., the inquisito-rial one) and attempt to incorporate practices from other systems—like pleabargaining—into their own legal systems.46

C. Procedural Powers

Finally, the adversarial and the inquisitorial differ at another level that Icall the “dimension of procedural power.”47 The main actors of the criminalprocess—judges, prosecutors, defense attorneys, defendants, police, etc.—have different quanta of procedural powers and responsibilities in each system.For example, from this perspective, the inquisitorial decision-maker, as anactive investigator, has more procedural power—e.g., to act sua sponte—thanthe adversarial decision maker.48 This also means that both the prosecutionand the defense in the inquisitorial system are comparatively less powerfulthan in the adversarial system. An example of this is the power that thedefense has in the adversarial system to do its own pre-trial investigation—apower generally not present in inquisitorial systems.49 The variation ofprocedural powers at the level of individual actors can also be seen at theinstitutional level in the relationships of power between the ofªce of theprosecution, the judiciary, the bar, the public defense ofªce, the police, etc.50

46. On the inºuence that U.S. legal education may have in processes of Americanization of European

legal rules and practices, see Wiegand, Americanization, supra note 1, at 139–40.47. This dimension of procedural power has also been relatively overlooked by comparative criminal

procedure analyses, and it is central not only to describing the differences between the adversarial and theinquisitorial systems, but also to identifying potential loci of resistance towards judicial reforms in adver-sarial and inquisitorial institutional settings.

48. This is clear if we compare the active judges of the inquisitorial system with the jury of the adver-sarial one. The inquisitorial judges are also more powerful than adversarial professional judges because oftheir power to decide which evidence is produced at trial and the order in which it is presented, as well asthrough their power to lead the interrogation of witnesses and expert witnesses. However, this last state-ment must be qualiªed. The adversarial judges have inherent powers—i.e., contempt powers—that theinquisitorial ones lack. In addition, since there is less hierarchical control over the decisions of the adver-sarial judges than the inquisitorial judges, the former also have more power in this respect.

49. Regarding France, see, for example, Valérie Dervieux, The French System, in European Criminal

Procedures 218, 250 (Mireille Delmas-Marty & J. R. Spencer eds., 2002).50. Though elaborating on this category more extensively is not necessary for the purposes of this Ar-

ticle, I also include within this dimension of procedural and institutional power not only the relationshipbetween the main permanent actors and institutions of the criminal justice system, but also the relation-


This dimension of procedural power also has a relationship of mutualinºuence with structures of meaning and internal dispositions. For instance,an inquisitorial structure of interpretation and meaning gives the judgebroad investigatory powers while giving more limited powers to the prose-cution and defense. At the same time, though, any attempt to change thisstructure of interpretation and meaning will usually generate a reaction bythe judges who protest against being disempowered through a new proce-dural structure of meaning.51

This leads to a very important point. Even though I have presented thedimensions of structure of interpretation and meaning, personal disposi-tions, and procedural and institutional powers as analytically different, inreality, they operate jointly and tend to reinforce, though also eventuallysubvert, one another.

D. Other Elements of the Systems

The interactions among these three dimensions have additional implica-tions. The adversarial and the inquisitorial procedural structures of interpre-tation and meaning are not only the “lenses” through which legal actors un-derstand and operate in reality. They also constitute two normative ordersthat indicate, to a certain extent, how cases should be handled, what tech-nologies should be used, how each of the actors of the system should behave,etc.

For instance, they determine to a certain degree how material and humanresources are structured and managed. Examples of this are the differentcase-management techniques that exist in each system. In the inquisitorialsystem, a written dossier is the backbone of the whole process and one of itsmain case-management tools, from the ªrst stage of the proceeding in whichthe police intervene, to the phase of appeals against the verdict.52 Con-

ships between the permanent professional actors and lay people. In the inquisitorial system, the power oflay people as decision-makers is minimal or entirely non-existent. In the adversarial system, it is muchmore substantial, at least in comparative terms. See, e.g., Robert A. Kagan, Adversarial Legalism

86–87 (2001).

51. The implementation of the 1998 Code of Criminal Procedure of the Province of Buenos Aires,Argentina, provides an example of this phenomenon. This code moved the responsibility of making thepre-trial investigation from judges to prosecutors. Consequently, a good number of judges resisted thisreduction of their procedural powers. This resistance was manifested, for instance, in the so-called “pho-tocopies war.” When requesting search warrants, investigating prosecutors chose to submit to judgescertiªed copies of the written dossier that contained the investigation rather than the originals as a meansof speeding up the proceedings. However, a number of judges took this as a symbolic issue about whowas really in charge of the investigation. Thus, they said they would not grant any search warrants untilthey received the original written dossiers. These controversies spoiled several investigations and becamesuch a serious issue that the Supreme Court of the province had to intervene. See, e.g., Rafael Saralegui,Demora judicial en San Isidro por una disputa, La Nación (Arg.), June 6, 2000, at 1, available at http://lanacion.com.ar/Archivo/Nota.asp?nota_id=19704 (last visited Nov. 25, 2003) (on ªle with the HarvardInternational Law Journal).

52. For a description of the role of the written dossier in inquisitorial systems, see Rudolf B. Schlesin-ger, Comparative Criminal Procedure: A Plea for Utilizing Foreign Experience, 26 Buff. L. Rev. 361, 365–67(1977).


versely, in the adversarial system, oral and public hearings play an importantrole in the management of cases—even in those that are bargained.53 In asimilar way—as we will see in more detail later—plea bargaining has beenan unknown case-management tool in inquisitorial systems until recently,54

but it has been allowed and widely used in Anglo-American jurisdictions.55

This does not mean, however, that the dimensions of procedural structureof meaning, internal dispositions, and procedural power, inºuence the dis-tribution of human and material resources, professional ethics, the internalsystem of incentives, etc., in a unidirectional way. These sets of factors aremutually inºuential. For example, the existence of a written dossier, a case-management tool that contains all of the evidence gathered during the pre-trial phase is necessary for the trial judge to behave as an active investigatorat trial.56 If the written dossier did not exist because, for instance, it wassuppressed through a legal reform, the trial judge could not behave in suchan active way; she could not organize the trial in advance, interrogate thewitnesses effectively, etc., and the parties would gain procedural powers ather expense.57 Provided this reform lasted, it could produce a change in theinternal dispositions of the judges, prosecutors, and defense attorney of theinquisitorial system, so that they start to behave in a different way: thejudges more passively, the parties more actively, etc. This change in criminalprocedure practices (the parole) would then produce a change in the inquisi-torial structure of interpretation and meaning (the langue) and in the dimen-sion of procedural power.

Taking into account these multi-directional relations between the threedimensions described and the material and human resources, case-managementtechniques, professional ethics, etc., will be important in analyzing why andhow the adversarial and inquisitorial criminal procedures change over time,including the changes produced by the transfer of legal ideas and institu-tions.

53. On the preference for oral production of witness testimony in open court in the United States, see

Fed. R. Crim. P. 26. On the importance of oral communication in English criminal procedure, seeDamaška, supra note 24, at 61.

54. For a classic analysis, see John H. Langbein, Land Without Plea Bargaining: How the Germans Do It,78 Mich. L. Rev. 204 (1979).

55. For historical analyses of how the practice of plea bargaining developed in U.S. jurisdictions, seeAlbert W. Alschuler, Plea Bargaining and Its History, 13 Law & Soc’y Rev. 211 (1979); George Fisher,Plea Bargaining’s Triumph, 109 Yale L.J. 857 (2000); Lawrence M. Friedman, Plea Bargaining in HistoricalPerspective, 13 Law & Soc’y Rev. 247 (1979); John H. Langbein, Understanding the Short History of PleaBargaining, 13 Law & Soc’y Rev. 261 (1979).

56. See, e.g., Mirjan Damaška, Aspectos globales de la reforma del proceso penal, in Reformas a la Justicia

Penal en las Américas (Fundación para el Debido Proceso Legal ed., 1999), available at http://www.dplf.org/Conference98/Damaska.pdf (last visited Oct. 18, 2003).

57. This was precisely the idea behind Article 431 of the Italian Criminal Procedure Code of 1989that eliminated trial judges’ full access to the written dossier collected during the pre-trial phase. SeeCodice di procedura penale, art. 431 (2002) (Italy) [C.p.p.]. For an analysis of this reform, seeGrande, supra note 32, at 243–44.


E. Points of Emphasis

Two ªnal points deserve emphasis. First, as the adversarial and the in-quisitorial categories exist within the concrete criminal procedure practicesof common and civil law jurisdictions, they are constantly challenged andsubjected to change. For instance, every time a particular U.S. judge activelyinterrogates witnesses (or allows juries to do it), there is a challenge to theadversarial structure of interpretation and meaning. Nevertheless, the factthat such challenges happen all the time does not mean that the adversarialand the inquisitorial systems do not continue to exist within these criminalprocedure practices. The overarching conception of the system may still behighly predominant within these practices even as we ªnd a number ofexceptions and challenges. In this sense, I do not believe these proceduralcultures—and culture in general—to be homogeneous. My empirical claim isthat the adversarial and the inquisitorial procedural cultures remain highlypredominant in Anglo-American and civil law jurisdictions of continentalEurope and Latin America respectively, even today, at least as concerns theformal proceedings through which guilt or innocence are determined.

Second, even if the adversarial or the inquisitorial categories have beenhighly predominant during the formal proceedings for determining guilt orinnocence, this does not mean that they are the only structures of interpreta-tion and meaning that exist respectively in the criminal justice systems ofAnglo-American and civil law jurisdictions.58 Despite the existence of these

58. Addressing this issue in detail is beyond the scope of this Article. But I would like to stress four

main points. First, there are local procedural differences that cannot be reduced to these global structuresof meaning. For instance, the Ofªce of the Prosecutor is part of the Executive Power in France, of thejudiciary in Italy, and an institution independent of the three traditional branches of government inArgentina. These different institutional positions have implications for what a prosecutor is understoodto be in these three countries—i.e., how independent they are from the Executive Power (for an analysisof the relationship between prosecutorial independence and the institutional position of the ofªce of theprosecutor in the Latin American context, see Philip B. Heymann, Should Latin American Prosecutors beIndependent of the Executive in Prosecuting Government Abuses?, 26 U. Miami Inter-Am. L. Rev. 535(1995))—and these different understandings cannot be reduced to the inquisitorial and adversarial struc-tures of meaning. (For an analysis of the consequences of these different institutional positions of prosecu-tors in France and Italy, see Carlo Guarnieri, Prosecution in Two Civil Law Countries: France and Italy, inComparing Legal Cultures 183 (David Nelken ed., 1997)). Second, in every concrete jurisdiction,there are generally other structures of interpretation and meaning in criminal procedure and, morebroadly, in the criminal justice system, that compete with or complement the adversarial and the inquisi-torial. For example, for most of the twentieth century in the United States, the sentencing phase of theproceedings has not been structured or mainly inºuenced by the adversarial system, but rather by a pa-ternalistic conception of the proceedings that considered the convicted person as a dangerous or sickperson who has to be rehabilitated. (For an analysis of the sentencing phase in the United States duringmost of this century and how it changed with the introduction of the Federal Sentencing Guidelines, see,for example, Kate Stith & José A. Cabranes, Judging Under the Federal Sentencing Guidelines, 91 Nw. U. L.

Rev. 1247 (1997)). Even in the phase of determination of guilt, where the adversarial system has pre-vailed for a long time in the United States, there are speciªc proceedings that are organized according toan inquisitorial structure of meaning, like the ones before the grand jury. On the inquisitorial characterof the grand jury, see, for example, Abraham S. Goldstein, Reºections on Two Models: Inquisitorial Themes inAmerican Criminal Procedure, 26 Stan. L. Rev. 1009, at 1020 (1974). For a proposal to introduce moreadversarial protections in grand jury proceedings, see Peter Arenella, Reforming the Federal Grand Jury andthe State Preliminary Hearing to Prevent Conviction without Adjudication, 78 Mich. L. Rev. 463 (1980).


other structures of interpretation and meaning that operate at the proce-dural, legal, and social levels, it would be a mistake to underestimate thegreat importance of the adversarial and inquisitorial ones. The adversarialand the inquisitorial have exerted a very substantial inºuence on criminalprocedure practices, norms, organizational arrangements, systems of incen-tives, etc., in Anglo-American and civil law jurisdictions, especially duringthe phase of determination of guilt or innocence. Even if speciªc analyseshave to pay attention to local particularities and to social, economic, andpolitical inºuences on the criminal process, it would be a mistake not to attendto these substantial global commonalities or differences. Using a metaphortaken from modern systems theory, one could say that the Anglo-Americanand civil law systems respectively share the same or a similar adversarial orinquisitorial program, even if this program co-exists with other legal andsocial programs in each particular jurisdiction.59 These programs have to beadequately conceptualized in order to analyze the similarities and differencesbetween the Anglo-American and civil law jurisdictions and the circulationof legal ideas and institutions between them, as well as to discuss the subtle-ties of the Americanization thesis.

III. The Content of the Dichotomy: Adversarial vs. Inquisitorial

Having explained the underpinnings of the adversarial and inquisitorialsystems as theoretical categories, there are two binaries that I will use in thisPart to explain the main differences between these systems. First, whereasthe adversarial system conceives of criminal procedure as a dispute betweenprosecution and defense before a passive umpire, the inquisitorial systemconceives criminal procedure as an ofªcial investigation carried out by ofªcialsof the state in order to determine the truth. Second, whereas the adversarialsystem’s decision-maker is a jury that divides its work with a professionaljudge, the inquisitorial system’s decision-maker is a professional judge or agroup of professional judges who do not share their responsibility with any

Third, there are other structures of interpretation and meaning regarding not only the criminal justicesystem, but also the legal system as a whole. If some of these structures can be reduced to global distinc-tions between the common and the civil law—i.e., legal reasoning in the former is more based onanalogical inferences and in the latter on deductive ones—there are other differences that cannot. Forexample, while some countries have a federal system, others do not, and this difference does not appear tooriginate in the adversarial and the inquisitorial systems. Finally, there are also other structures of inter-pretation and meaning (political, economic, religious, media, etc.) that exceed the legal sphere but thatmay affect how criminal procedure and criminal justice actors are understood. For instance, the prelimi-nary investigation judge—juge d’instruction in Francophone countries, juez de instrucción in Latin Americaand Spain—is on many occasions represented in the French social imagination as the small ofªcial who“makes the truth stand out” by ªghting “against the powerful, the strength of ink and law against that ofmoney and power,” and, at the same time, is an emblematic ªgure of the provinces in their struggle withParis. Garapon, supra note 37 at 497–98. This is a social representation of the preliminary investigationjudge that is typically not found in other inquisitorial systems, such as Argentina.

59. See Gunther Teubner, The Two Faces of Janus: Rethinking Legal Pluralism, 13 Cardozo L. Rev.

1443, 1455 (1992).


other body.60 These binaries explain some of the critical differences betweenthe adversarial system of the United States and the inquisitorial systems ofGermany, Italy, Argentina, and France.

Attempting to deªne the adversarial and the inquisitorial systems for thepurposes of comparative law is a difªcult task for at least four different rea-sons.61 First, differences between the criminal procedures of the commonand civil law traditions can be traced back to the thirteenth century whenEngland and Europe developed different systems to replace the then pre-vailing practices that had been in place since the fall of the Western RomanEmpire.62 These differences have evolved over time, and explaining theseseparate evolutions is beyond the scope of this Article.63

Second, as a consequence of phenomena like colonization, civilization, andmodernization, these two traditions have each expanded to include a sub-stantial number of countries with varying individual legal rules and prac-tices, thus complicating efforts to capture the differences and similaritiesbetween the systems as a whole.64

Third, both in Anglo-American and civil law jurisdictions, the expres-sions “adversarial” (or “accusatorial”)65 and “inquisitorial” are fraught with

60. As I explain later, professional judges may share their responsibility with lay people in mixed

courts. However, even in this case, professional judges still usually retain control of the decision-makingprocess, given that they are the professional and permanent actors. Furthermore, there is only one deci-sion-making body composed by professional judges and lay people, instead of two—the judge and thejury—as in the adversarial system.

61. The expressions “accusatorial” and “inquisitorial” were already in use during the twelfth centuryin Europe “to distinguish a process that required the impetus of a private complainant to get under way(processus per accusationem) from a process that could be launched in his absence (processus per inquisitionem).”Damaška, supra note 24, at 3. The modern use of these expressions, that conceives of the accusatorial andthe inquisitorial not only as two different ways of initiating procedures, but also as two comprehensiveprocedural systems, was likely to have developed during the nineteenth century. The ªrst such usage ofwhich I am aware is Faustin Hélie, 5 Traité de l’Instruction Criminelle ou Théorie du Code

d’Instruction Criminelle 47–65 (1853).62. For a description of the system of ordeals, trial by combat, and oaths prevalent in Europe prior to

the thirteenth century and an explanation of why this disappeared, see Robert Bartlett, Trial by

Fire and Water (1986).

63. A classic account is Adhemar Esmein, Histoire de la Procédure Criminelle en France

(1882). There is an English edition of this work: Adhemar Esmein, A History of Continental

Criminal Procedure (John Simpson trans., 1968) (1882). For contemporary accounts of these devel-opments, see Franco Cordero, Procedura Penale 16–101 (2d ed. 1993); John H. Langbein,

Prosecuting Crime in the Renaissance (1974); John H. Langbein, The Origins of Adversary

Criminal Trial (2003); John H. Langbein, Torture and the Law of Proof (1977); Julio B.J.

Maier, 1 Derecho Procesal Penal § 5 (2d ed. 1996); Jean-Pierre Royer, Histoire de la justice

en France: de la monarchie absolue à la République (2d ed. 1996). I am not referring to “evolu-tion” as progress or movement towards perfection, but simply as change over time. For an interestinganalysis regarding the use of this term in system theory, see Gunther Teubner, Legal Irritants: Good Faithin British Law, or How Unifying Law Ends Up in New Divergences, 61 Mod. L. Rev. 11, 15–16, 27–32(1998).

64. On the expansion of the common law throughout the world, see Konrad Zweigert & Hein

Kötz, Introduction to Comparative Law 218–37 (3d ed. 1998). For a description of how the in-quisitorial system was imposed and developed in Latin America, see Maier, supra note 63, § 5(D)(8). Foran analysis of these developments in a number of African, Asian and Inter-American countries, see Jean

Pradel, Droit Pénal Comparé 186–201 (1995).65. Abraham S. Goldstein tried to establish a distinction between the terms “adversarial”—as a way of


political and cultural connotations; for instance, the adversarial tradition isusually linked to liberal or democratic conceptions while the inquisitorialtradition is linked to authoritarian conceptions of criminal procedure.66 Thishas led to what could be described as a rhetorical struggle for the appropria-tion of these terms, that in turn has multiplied the terms’ different uses.67 Infact, as a consequence of these connotations, “adversarial” and “inquisitorial”have been central terms or “ºoating signiªers” through which the actors ofthe Anglo-American and the civil law systems have deªned and differenti-ated their own identity, both from the identity of other traditions as well asfrom their own past.68

Finally, even if it is possible to identify historical differences between thecriminal procedures of the common and the civil law that have lasted untiltoday, it is not always possible to reduce these differences to a common ex-planatory principle. For example, plea bargains and detailed rules of evi-

ªnding facts and implementing norms—and “accusatorial”—that would include not only adversary trialprocedures but also a conception of the state as being neutral in disputes. However, in this Article, Iconsider these terms as equivalent. See Abraham S. Goldstein, Reºections on Two Models: Inquisitorial Themesin American Criminal Procedure, 26 Stan. L. Rev. 1009, 1017 (1974).

66. For Argentina, see, for example, Maier, supra note 63, § 5(H)(1)(a)–(b). For France, see, for exam-ple, Jean Pradel, Inquisitoire-Accusatoire: une redoutable complexité, 68 Int’l Rev. Penal L. 213, 215 (1997).For Germany, see, for example, Claus Roxin, Strafverfahrensrecht § 2, at 9–11 (25th ed. 1998).For Italy, see, for example, Luigi Ferrajoli, Diritto e Ragione: Teoria del Garantismo Penale

576, 655 n.84 (1989). For the United States, see, for example, Miranda v. Arizona, 384 U.S. 436, 442–43, 459–60 (1966).

67. See Langer, supra note 23, at 102–14.68. In the United States, the word “adversarial” has been used in laudatory terms in reference to U.S.

criminal process, which had its earliest origins in England in the struggles for rights against the authori-tarian monarchy, struggles that continued in the Colonies and that ªnally found their way into the Bill ofRights. In this construction of the adversarial, the inquisitorial refers to the contemporary criminal pro-cedures of continental Europe that would still be considered authoritarian—i.e., eliciting confessions in acoercive way. For an example of this construction of the adversarial and the inquisitorial in the UnitedStates, see Miranda, 384 U.S. at 442–43, 459–60. (For another use of the terms in the United States thatis not presented in a laudatory way and is focused on the role of the judge in each of these systems, see,for example, McNeil v. Wisconsin, 501 U.S. 171, 181 n.2 (1991).) In continental Europe, one of themost widespread uses of the accusatorial and the inquisitorial for deªning the identity of its moderncriminal procedure has been the following. The accusatorial is used to refer to the modern Anglo-American criminal procedures—and the ones that prevailed in continental Europe from the fall of theWestern Roman Empire until the thirteenth century—that are usually considered inefªcient in lawenforcement terms; the inquisitorial refers to the criminal procedures that prevailed in continentalEurope from the thirteenth to the nineteenth centuries, that are usually characterized as authoritarian;and the modern continental European criminal procedures would constitute a mixed system that wouldcombine the best of the two other systems. For an example of this use in early twentieth-century France,see René Garraud, 1 Traité Théorique et Pratique d’Instruction Criminelle et de Pro-

cédure Pénale 10–22 (1907). Hélie uses the terms in a similar way, Hélie, supra note 61, at 47–65.For a contemporary use of these terms in Latin America, see Maier, supra note 63, § 5(H)(1)(a)–(c), at443–54, though this author is very critical of the mixed system. I am not aware that anyone else hasanalyzed the inquisitorial and the adversarial as central terms or signiªers through which the identities ofthe actors of the common and the civil law traditions have been deªned. Understanding these processes ofdeªning identity could open new doors for comparative criminal procedure analyses. For instance, theymay be useful in understanding why continental European and Latin American actors have been inter-ested in adopting institutions and ideas from the Anglo-American system, rather than the other wayround. For a recent analysis of legal traditions as identities, see H. Patrick Glenn, Legal Traditions

of the World (2000).


dence have traditionally been both associated with the Anglo-Americancriminal process and considered non-existent in most civil law countries.69

Nevertheless, the historical and analytical explanations of these two featurescannot be reduced to a single source. Plea bargains seem to have their originin a criminal procedure system understood as a dispute between parties, whereasdetailed rules of evidence seem to be mainly linked to the use of a bifurcatedcourt where one organ—the judge—decides what evidence can be intro-duced at trial, and another organ—the jury—decides the guilt or innocenceof the defendant.70

In this Article, I address such difªculties as follows. First, regarding theproblem of change in the systems over time, I focus on contemporary crimi-nal procedures of civil law systems before they introduced the speciªc adver-sarial reforms analyzed in the next Parts. Second, regarding the plurality ofjurisdictions that could be included in the adversarial and inquisitorial cate-gories, I have chosen to focus on ªve jurisdictions. When I refer to the ad-versarial system, I am referring to U.S. jurisdictions, and when I refer to theinquisitorial system, I am referring to the criminal procedures of Germany,Italy, Argentina, and France. Third, concerning the political connotations ofthe terms “adversarial” and “inquisitorial,” it is necessary to stress that I usethem only as descriptive tools for the purposes of comparative law. Finally,rather than attempting to explain the differences between the systemsthrough one explanatory principle, I instead identify two models withineach system as the bases for comparison.71

The ªrst pair of models for comparison are the model of the dispute (a featureof the adversarial system) and the model of the ofªcial investigation (a feature ofthe inquisitorial system).72 According to the ªrst model, the criminal proce-

69. See, e.g., Langbein, supra note 39, at 68–71.70. Id. at 70–71. For a study of why the law of evidence in Anglo-American jurisdictions exists in its

current form, and how it relates to other features of the Anglo-American system, see Mirjan R.

Damaška, Evidence Law Adrift (1997). Damaška does not believe that the existence of the jury hasplayed as central a role as is usually thought. Id. at 26–57.

71. Given my conceptualization of the adversarial and the inquisitorial systems as procedural cultures,there is a ªfth problem in giving content to both categories. As explained in Part II, these cultures arenot completely homogeneous, and within the criminal procedure practices of the United States, Argen-tina, France, Germany and Italy, one can identify different visions about what characterizes each of theselegal cultures. Thus, giving a particular content to each of these legal cultures may silence or suppresssome of these alternate visions and may contribute to make prevail one vision over others. The ªrst wayto address this problem is by recognizing it as such, in other words, by being self-conscious and self-reºexive about it. In addition, I think that the adversarial and the inquisitorial systems as I deªne themin this Part have been overwhelmingly predominant in these ªve jurisdictions—at least during the phaseof determination of guilt and innocence. Therefore, I will concentrate in this Article on these predomi-nant conceptions of the adversarial and inquisitorial because they are the most useful in explaining howmost legal actors in each of these jurisdictions understand criminal procedure, and thus how plea bar-gaining was transformed when adopted by our four civil law countries. Works focused on the dissidentcultural voices in each of these jurisdictions and on the heterogeneity of the adversarial and inquisitorialcultures would be very valuable, but these are part of a different endeavor than the one undertaken in thisArticle.

72. The idea of the dispute and the investigation has been used for comparative law purposes for along time. See, e.g., Hélie, supra note 61, at 53.


dure is understood as a dispute or a contest between two parties, prosecutionand defense, before a passive decision-maker. The dispute centers around theprosecution’s attempt to prove beyond a reasonable doubt that the defendantcommitted the offense of which he or she has been accused. If the prosecu-tion succeeds, then the prosecution wins; if it fails, the defendant wins.

Many characteristic features of Anglo-American criminal procedure can beexplained through this model. For instance, broad prosecutorial discretionªts within this model because the prosecution, as one of the parties andowners of the dispute, may not believe that there is controversy in a par-ticular case or may decide that the controversy is not worth pursuing and isthus not obligated to pursue it.73 Guilty pleas ªt into this model because thedefense, as the other party to the dispute, may concede that the other partyis correct and thus resolve the dispute; the determination of guilt or inno-cence ends74 and the case passes to the sentencing stage. If criminal proce-dure is seen as a dispute, then structuring the process as two competing andclashing cases also ªts within this conception. Thus, in Anglo-Americanjurisdictions, each party to the dispute does its own separate pre-trial inves-tigation,75 even if the parties must disclose part of the information theygather to the other party through discovery rules and proceedings.76 Thetrial is divided into a case for the prosecution and a case for the defense; theparties usually decide in what order the evidence will be presented; the wit-nesses and expert witnesses belong to the prosecution or the defense77 and

73. For analyses of the broad discretion of U.S. prosecutors, see, for example, Norman Abrams, Prose-

cution: Prosecutorial Discretion, in 3 Encyclopedia of Crime and Justice 1272 (Sandorf H. Kadish ed.,1983); Richard S. Frase, The Decision to File Federal Criminal Charges: A Quantitative Study of ProsecutorialDiscretion, 47 U. Chi. L. Rev. 246 (1980); Peter Krug, Prosecutorial Discretion and its Limits, 50 Am. J.

Comp. L. 643 (2002); Wayne R. LaFave, The Prosecutor’s Discretion in the United States, 18 Am. J. Comp. L.

532 (1970); Robert L. Misner, Recasting Prosecutorial Discretion, 86 J. Crim. L. & Criminology 717

(1996); James Vorenberg, Decent Restraint of Prosecutorial Power, 94 Harv. L. Rev. 1521 (1981). For classi-cal analyses of how to address the problem of discretion in the criminal justice system, regarding not onlyprosecutorial practices but also the practices of other legal actors, see Kenneth Culp Davis, Discre-

tionary Justice: A Preliminary Inquiry (1969); Arthur Rosett, Discretion, Severity and Legality inCriminal Justice, 46 S. Cal. L. Rev. 12 (1972).

74. See, e.g., Fed. R. Crim. P. 11(c)(4).75. On the duty of the defense attorney in the United States to make reasonable investigations, or to

make reasonable decisions that particular investigations are unnecessary, see Wiggins v. Smith, 123 S. Ct.2527 (2003); Strickland v. Washington, 466 U.S. 668, 691 (1984).

76. In fact, until the promulgation of Federal Rule of Criminal Procedure 16 in 1946, no right to dis-covery existed in federal litigation. See Frank W. Miller et al., Criminal Justice Administration:

Cases and Materials 753 (5th ed. 2000). Today, discovery rules are grounded in the Constitutionunder the Brady doctrine, Strickler v. Greene, 527 U.S. 263 (1999); Kyles v. Whitley, 514 U.S. 419(1995); United States v. Bagley, 473 U.S. 667 (1985); United States v. Agurs, 427 U.S. 97 (1976); Gigliov. United States, 405 U.S. 150 (1972); Brady v. Maryland, 373 U.S. 83 (1963); in federal statutes, see,e.g., 18 U.S.C. § 3500 (1970), in the Federal Rules of Criminal Procedure, see, e.g., Fed. R. Crim. P. 12.1,12.2, 16, and 26.2; and in the supervisory powers of the courts.

77. On the coaching that lawyers give their witnesses in the United States, see, for example, William

T. Pizzi, Trials without Truth 21–22 (1999). But see Fed. R. Evid. 614(a) (establishing that the courtmay, on its own motion, call a witness); Fed. R. Evid. 706(a) (stating that the court may also appointexpert witnesses of its own selection). These powers, however, are seldom used in criminal trials. See, e.g.,United States v. Ostrer, 422 F. Supp. 93 (S.D.N.Y. 1976) (stating that although the court has discretion-


the questioning of witnesses is developed as a dispute between two parties,with direct and cross-examination, as well as redirect.78 Plea bargains canalso be explained through the dispute model because it is natural in any dis-pute that the parties can negotiate a resolution.79

In the model of the ofªcial investigation, which corresponds to the in-quisitorial system, criminal procedure is conceptualized as an inquiry madeby one or more ofªcials of the state in order to determine whether a crimewas committed and whether the defendant committed it.80 Numerous ele-ments of the criminal procedure in civil law countries can be explainedthrough the use of this model.81 Compulsory prosecution is necessary withinthis model because the proceedings are an investigation to determine thetruth, and thus, the case can only be dismissed when there is no evidencethat an offense has been committed or that the defendant committed it.82

The concept of the guilty plea does not exist in this model because, whilethe admission of guilt by the defendant can be a very important element ofproof, it does not necessarily provide a complete version of the truth, whichis for the judge to decide.83 Further, there are no plea bargains,84 not onlybecause there are no guilty pleas but also because the truth cannot be nego-tiated or compromised. Finally, the whole procedure is structured and un-

ary power to call a court witness, this power is rarely invoked). For an analysis of the obstacles that judgeswould face in becoming more active players at trial, see Marvin E. Frankel, The Search for Truth: An Um-pireal View, 123 U. Pa. L. Rev. 1031, 1041–45 (1975).

78. See, e.g., Fed. R. Evid. 611. Federal Rule of Evidence 614(b) establishes that “the court may inter-rogate witnesses, whether called by itself or by a party.” But this power is also infrequently used incriminal trials. See Bradley, supra note 43, at 421.

79. See, e.g., Brady, 397 U.S. 742 (approving, for the ªrst time, the constitutionality of plea bargain-ing). On the history of plea bargaining in the United States, see the references at supra note 55.

80. For instance, regarding the determination of truth being the goal of criminal procedure in France,see Michèle-Laure Rassat, Traité de Procédure Pénale 297 (2001).

81. Recall that this description refers to the criminal procedures of Argentina, France, Germany, andItaly, as they were before the reforms that I will analyze in this Article. This is particularly important inthe case of Italy, which changed many of these features when it moved in the direction of the model ofthe dispute by introducing its Criminal Procedure Code of 1989. However, even if the description refersto their pre-reform criminal procedures, most of these features are still present today in most of thesecivil law countries. When I cite rules, I am referring to rules that are still valid today.

82. There are differences in the way the four main civil law jurisdictions discussed in this Articleregulate this issue. For jurisdictions establishing rules of compulsory prosecution, see Código Penal[Cód. Pen.] [Penal Code] art. 59.4 (Arg.) (for Argentina); Costituzione della Repubblica Italiana [Cost.]

art. 112 (Italy). Germany established a rule of compulsory prosecution as the general rule, although thereare exceptions through its opportunity principle. §§ 152–54a Strafprozessordnung [StPO] [criminalprocedure statute] (F.R.G.). In France, the prosecutor has discretion about whether to bring chargesinitially. Code de procédure pénale [C. pr. pén.] [Criminal procedure code] art. 40 (Fr.). But if theFrench prosecutor decides to bring charges, he/she cannot dismiss the charges without the acquiescence ofthe court. See Rassat, supra note 80, at 452–53). (I will not analyze here the powers of the victim re-garding this issue). As there is no system that can possibly prosecute all criminal offenses, the rule ofcompulsory prosecution or limited prosecutorial discretion has been described as a myth. For a debate onthis issue in the United States, see Abraham S. Goldstein and Martin Marcus, The Myth of Judicial Super-vision in Three “Inquisitorial” Systems: France, Italy, and Germany, 87 Yale L.J. 240 (1977); John H. Lang-bein & Lloyd L. Weinreb, Continental Criminal Procedure: “Myth” and Reality, 87 Yale L.J. 1549 (1978).

83. See Moskovitz, supra note 38, at 1153. See also Langbein, supra note 39, at 73–74.84. See, e.g., Langbein, supra note 52.


derstood as a unitary investigation.85 Thus, there is only one pre-trial inves-tigation, the ofªcial one;86 at trial, there is no case for the prosecution or thedefense, only the case of the court; the court decides in what order the evi-dence is presented at trial;87 the witnesses and expert witnesses do not corre-spond to the parties but to the court;88 and the interrogations are initiatedand directed by the court and not by the parties.89

The relationships between the actors of the procedure in these two modelscould schematically be represented in the following way:

Model of the Dispute

Decision-Maker (Judge or Jury)

Prosecution Defendant/Defense Attorney

Police Private Investigator

Model of the Ofªcial Investigation

Prosecution Judge (Decision-maker and investigator)

Police90

Defendant/ Defense Attorney

In the model of the dispute, the prosecution and the defense are at thesame level, both in the sense that they have relatively equal procedural pow-ers and that they both are understood as parties that have an interest at stake

85. Regarding France, see, for example, Dervieux, supra note 49, at 250.86. See id.87. See, e.g., Código Procesal Penal de la Nación [Cód. Proc. Pen.] [Criminal Procedure Code] art.

356 (Arg.). Regarding Germany, see §§ 155 Nr. 2, 244 Nr. 2 StPO (F.R.G.). See also Roxin, supra note66, § 15, at 95.

88. France is an exception because, during trial, the witnesses belong to the parties. See, e.g., Rassat,

supra note 80, at 405–06. However, the witnesses cannot be coached, and it is still the presiding judgewho questions witnesses acting sua sponte or on the request of the parties. See C. pr. pén. arts. 312, 332,454, 536 (Fr.). From January 1, 2001, questions can also be put to the witnesses directly by the parties.See Dervieux, supra note 49, at 258–59.

89. Regarding Argentina, see Cód. Proc. Pen. art. 389 (Arg.). Regarding Germany, see §§ 238 Nr.1, 240 StPO (F.R.G.). Section 239 of StPO states that interrogatories at trial shall be developed throughdirect and cross-examination if the prosecution and the defense request it. However, this is rarely applied.See Roxin, supra note 66, § 42, at 343.

90. Depending on the civil law jurisdiction, the police may work under the supervision of the prose-cutor or the judge—generally depending on who is in charge of the pre-trial investigation, or both.


in the case.91 Each of them has its own investigator—the police in the caseof the prosecution, a private investigator in the case of the defense. The deci-sion-maker is above them but can only make decisions about the case solong as both parties believe there is a dispute.

There is also a prosecutor in the model of the ofªcial investigation. He orshe is not seen as a party but rather as another ofªcial or magistrate of thestate whose role is to determine the truth.92 There is a sort of division oflabor between the prosecutor and the judge: the ªrst requests the investiga-tion of facts, production of evidence, and the application of the law, whilethe second investigates, produces the evidence, and applies the law. But bothare essentially the same: impartial ofªcials of the state whose role is to inves-tigate the truth.93 In the schematic provided, the defendant is representedbelow the judge and the prosecutor not because she does not have rights, butrather because she has an interest at stake in the resolution of the case, andbecause she is not only a subject of rights but also a target of investigation.94

There is no private investigator for the defense because there is only oneofªcial investigation, conducted by the judge, the prosecutor, and the police.If the defense wants certain evidence to be produced, it must request thisevidence from the prosecutor or the judge.95

The second pair of models that I employ to capture the differences be-tween the criminal procedures of the common and the civil law are what

91. In the United States, it is usually said that the role of the prosecutor is not only to look for a con-

viction, but also for justice. Nevertheless, in comparative terms, the U.S. prosecutors usually act as par-ties with an interest at stake in the case. For instance, they usually consider that they win a case when thedefendant is convicted and lose a case when the defendant is acquitted or the case is dismissed withprejudice. See Pizzi, supra note 35, at 1349.

92. See Weigend, supra note 35, at 1233–34.93. See id.94. Many commentators have said that while in the adversarial system the defendant is a subject of

rights, in the inquisitorial system he is an object of investigation. This was probably true for a long time,but after World War II—and even earlier in some jurisdictions—most inquisitorial countries beganconsidering the defendant as a subject of rights, both at the rule level—constitutions, human rightstreaties, criminal procedure codes—and at the law in action level. Thus, in most civil law jurisdictionstoday, the defendant is presumed innocent, has a right against compelled self-incrimination, a right toassistance of counsel, etc. See, e.g., Gordon Van Kessel, European Perspectives on the Accused as a Source ofTestimonial Evidence, 100 W. Va. L. Rev. 799 (1998). This does not mean, of course, that all the rightsenunciated in rules are respected in practice. The point that I am trying to make is that the concept ofrights seems to have lost part of its old heuristic value in establishing clear distinctions between adver-sarial and inquisitorial systems. This is why I do not mention it as one of the differences between the twosystems, either in the opposition between the model of the dispute and model of the ofªcial investiga-tion, or in the opposition between the coordinate and hierarchical models. In a similar way, the conceptsof public vs. private prosecution had a heuristic value in the past because the Anglo-American jurisdic-tions developed a system of public prosecution much more recently than civil law jurisdictions. Today,this distinction no longer has a heuristic value in comparative analysis. Regarding the Dutch inºuence onthe development of the public prosecutor in the United States, see Albert J. Reiss, Jr., Public Prosecutorsand Criminal Prosecution in the United States of America, 1975 Jurid. Rev. 1 (1975). Regarding the Frenchinºuence on the same development, see André Fournier, Code de Procédure Criminelle de

l’État de New York 9 (1893).

95. See, e.g., Dervieux, supra note 49, at 250.


Professor Damaška calls the coordinate and hierarchical models.96 According toProfessor Damaška, in the coordinate model, authority is exercised by laydecision-makers in relatively horizontal relationships of power amongstthemselves who apply community standards in their decisions.97 In the hier-archical model, authority is exercised by professional legal decision-makers whoserelationships are hierarchal and who apply technical standards in their deci-sions.98

I employ these models in slightly modiªed forms. In the coordinate model, afeature of the adversarial process, I include lay participation through thejury,99 less hierarchical control,100 a preference for oral production of the evi-dence at trials,101 a bifurcated court composed of a lay organ (the jury) and aprofessional one (the judge), the absence of a mandatory justiªcation of theverdict,102 and detailed rules of evidence that ªlter the elements of proofallowed at trial and that guide the evaluation of the evidence that is ªnallyintroduced.103

In the hierarchical model—part of the inquisitorial process—I includeprofessional decision-makers,104 greater hierarchical control,105 an acceptanceof the elements of proof gathered during the pre-trial investigation and

96. See Damaška, supra note 24, at 16–71.97. See id. at 17.98. See id.99. U.S. Const. amend. VI; Duncan v. Louisiana, 391 U.S. 145 (1968).100. See Kepner v. United States, 195 U.S. 100 (1904) (afªrming that the prosecution cannot appeal

acquittals in the United States).101. See, e.g., Fed. R. Crim. P. 26, and Fed. R. Evid. 802.102. See Fed. R. Crim. P. 31.103. See, e.g., Fed. R. Evid.

104. The province of Cordoba in Argentina, France, Germany, and Italy have mixed courts—com-prised of professional and lay judges—to try the most serious criminal cases. But in these mixed courts,the professional judges are still the most inºuential decision-makers given their legal knowledge andexperience, and because professional and lay judges deliberate together. For an analysis of mixed courts,see John H. Langbein, Mixed Court and Jury Court: Could the Continental Alternative Fill the AmericanNeed?, 1981 Am. B. Found. Res. J. 195 (1981). A substantial number of the inquisitorial systems ofcontinental Europe tried to import the jury during the nineteenth century. The precursor was France,who set up the jury twice, in 1791 and 1808. Other countries, like Germany, followed its example. Nev-ertheless, today most inquisitorial systems have replaced the jury with mixed courts or do not have anykind of lay participation in the criminal justice system. (As I will mention later, two exceptions are Rus-sia and Spain, who introduced the jury into their respective systems in 1993 and 1995.) For accounts ofthe introduction of the jury by France and Germany during the nineteenth century, see the articles in-cluded in The Trial by Jury in England, France, Germany 1700–1900 (Antonio Pado Schioppaed., 1987). For an explanation of why France and Germany ªnally rejected this institution and adoptedmixed courts, see Otto Kahn-Freund, On Uses and Misuses of Comparative Law, in Selected Writings

294, 310 (1978). For an analysis of the turn against the jury court on the part of liberal and conservativelegal actors alike after about 1900, see Benjamin C. Hett, Death in Tiergarten and Other Stories: Murderand Criminal Justice in Berlin, 1891–1933, 343–45 (2001) (unpublished Ph.D. dissertation, HarvardUniversity) (on ªle with the Harvard International Law Journal). For an analysis of the history of the juryin Germany and its link to contemporary debates on political philosophy, see Markus Dirk Dubber, TheGerman Jury and the Metaphysical Volk: From Romantic Idealism to Nazi Ideology, 43 Am. J. Comp. L. 227

(1995).

105. See, e.g., Langbein, supra note 39, at 84–85.


documented in a written dossier,106 a unitary court that decides what evi-dence is admitted at trial and determines the verdict and sentence,107 thewritten justiªcation of that verdict,108 and few rules of evidence to ªlterwhat elements of proof are introduced at trial and how they must be evalu-ated.109

This description of the hierarchical and the coordinate models is, obvi-ously, short and schematic. Nevertheless, a more detailed description of bothcategories is not necessary because I make limited use of them in this Arti-cle.110 As will be clear later, the central categories to analyze the transfer ofplea bargaining from the adversarial to the inquisitorial system will be themodels of the dispute and of the ofªcial investigation.

IV. Recent Reforms in the Inquisitorial System and the

Americanization Thesis

If the Americanization thesis is to hold any weight, even in its weak form,some elaboration of the inºuence of the adversarial system on civil law sys-tems is necessary. This Part describes such inºuences in the realm of crimi-nal procedure. In the last three decades, and depending on the region andjurisdiction, a signiªcant number of civil law countries introduced adver-sarial reforms into their inquisitorial criminal procedures, which go both inthe direction of the dispute model and the direction of the coordinatemodel.111

106. Contemporary civil law countries rely on oral production of evidence at trial much more than in

the past. For instance, the civil law principle of immediacy establishes that the decision-makers, the prosecu-tion, and the defense must be present during the production of evidence that the decision-makers willevaluate to issue their verdict and sentence. See, e.g., Maier, supra note 63, § 8(D)(3)(c), at 877–78. How-ever, the elements of proof collected in the written dossier continue to have an inºuence on trial judges—who in most jurisdictions do not participate in pre-trial proceedings and only know of those elements ofproof because of the written records. For an analysis of this issue in Germany, see Mirjan Damaška, OfHearsay and Its Analogues, 76 Minn. L. Rev. 425, 449–52 (1992).

107. See, e.g., Cód. Proc. Pen. arts. 356, 398 (Arg.).108. See, e.g., Cost. art. 111.6 (which was in place even before the reform of November 1999).109. See, e.g., Roger Merle & André Vitu, II Traité de Droit Criminel 193–94 (5th ed. 2001).110. See infra Parts IV and VIII.111. In the last thirty years, there has been a tendency on the part of Anglo-American legal academia

to propose reforms inspired by the civil law tradition. However, despite the quality of these works andthe fact that they were written by distinguished scholars, this tendency has been quite marginal withinthe U.S. criminal justice system and has not resulted in concrete reforms. Among the academic worksthat have been inspired by civil law criminal procedure, see Kenneth Culp Davis, supra note 73, at191–95; Robert A. Kagan, Adversarial Legalism 86–87 (2001); J.R.S. Spencer, French and EnglishCriminal Procedure, in The Gradual Convergence 33–45 (B. S. Markesinis ed., 1994); Pizzi, supranote 77; Lloyd L. Weinreb, Denial of Justice (1977); Albert W. Alschuler, Implementing the CriminalDefendant’s Right to Trial: Alternatives to the Plea Bargaining System, 50 U. Chi. L. Rev. 931 (1983); MarkusDirk Dubber, American Plea Bargains, German Lay Judges, and the Crisis of Criminal Procedure, 49 Stan. L.

Rev. 547 (1997); Richard S. Frase, Comparative Criminal Justice as a Guide to American Law Reform: HowDo the French Do It, How Can We Find Out, and Why Should We Care?, 78 Cal. L. Rev. 539 (1990); Rich-ard S. Frase & Thomas Weigend, German Criminal Justice as a Guide to American Law Reform: Similar Prob-lems, Better Solutions?, 18 B.C. Int’l & Comp. L. Rev. 317 (1995); Krug, supra note 73; Langbein, supranote 54; Langbein, supra note 104; Gordon Van Kessel, Adversary Excesses in the American Criminal Trial,67 Notre Dame L. Rev. 403 (1992). For an analysis of why there have not been reforms inspired by the


In moving toward the model of the dispute, a substantial number of con-tinental European and Latin American countries have tried to make thejudge a more passive actor by increasingly relying on the prosecution andthe defense for fact-ªnding and conducting proceedings. For instance, asigniªcant number of countries have eliminated the preliminary investiga-tion judge112—a symbol of the inquisitorial process—and have attempted toreplace this ªgure with a pre-trial investigation developed by the prosecu-tion and, in some cases, by the defense.113 During the pre-trial phase, therole of the judge is now largely limited to decisions about issuing search-and-seizure warrants, arrests, and pre-trial detentions and bails, much likethe Anglo-American model. Germany (1975),114 Italy (1989),115 Guatemala(1994),116 the provinces of Buenos Aires and Cordoba in Argentina (1998),117

Costa Rica (1998),118 Venezuela (1999),119 and Chile (2000–2003),120 amongothers, have moved in this direction.

Moreover, a number of jurisdictions have also altered the relative roles ofthe prosecution and defense during trial. For example, Italy has introduceddirect and cross-examination into its trial proceedings121 and has allowedboth the prosecution and defense to call witnesses.122 Italy has also struc-tured its trials by dividing them into a case for the prosecution and a case for

inquisitorial system in the U.S. adversarial system in recent years, see John H. Langbein, The Inºuence ofComparative Procedure in the United States, 43 Am. J. Comp. L. 545 (1995); John H. Langbein, The Inºuenceof the German Emigrés on American Law: The Curious Case of Civil and Criminal Procedure, in Der Einºuss

deutscher Emigraten auf die Rechtsenwicklung in den USA und in Deutschland 321 (Mar-cus Lutter et al. eds., 1993).

112. This ªgure is known as the Juge d’Instruction in France, the Untersuchungsrichter in Germany, theGiudice Istruttore in Italy, and the Juez de Instrucción in Latin America.

113. For instance, Costa Rica has put the prosecutor in charge of the pre-trial investigation. SeeCódigo Procesal Penal de Costa Rica [Cód. Proc. Pen.] [Code of Criminal Procedure] arts. 290–91(Costa Rica). However, the defense is not supposed to do its own pre-trial investigation; if the defensewants that certain elements of proof be gathered during the pre-trial phase, it has to request this gather-ing from the prosecution. See Cód. Proc. Pen. art. 292 (Costa Rica). But in the new Italian criminalprocedure in which the prosecution is in charge of the pre-trial investigation, the defense can gather itsown elements of proof during the pre-trial phase. See infra notes 164 and 225.

114. See Roxin, supra note 66, § 72, at 533.115. See, e.g., Giuliano Vassalli, Premessa, in Istituto poligraªco e zecca dello Stato, Progetto

Preliminare del Codice di Procedura Penale 5, at 6–7 (1988); Ennio Amodio & Eugenio Selvaggi,An Accusatorial System in a Civil Law Country: The 1988 Italian Code of Criminal Procedure, 62 Temp. L.

Rev. 1211, 1218–19 (1989).116. See Luis Rodolfo Ramírez García & Miguel Ángel Urbina, Guatemala, in Las Reformas

Procesales Penales en América Latina [hereinafter Las Reformas] 443, at 458–59, 471–72 (JulioB.J. Maier et al. eds., 2000).

117. See Fabricio Guariglia & Eduardo Bertoni, Argentina, in Las Reformas, supra note 116, at 35,53–54.

118. See Daniel González Álvarez, Costa Rica, in Las Reformas, supra note 116, at 269, 287–88.119. See Sergio Brown Cellino, Venezuela, in Las Reformas, supra note 116, at 769, 803.120. On the difªculties of introducing these kinds of reforms in an inquisitorial jurisdiction like

Chile, see Carlos Rodrigo de la Barra Cousino, Adversarial vs. Inquisitorial Systems: The Rule of Law andProspects for Criminal Procedure Reform in Chile, 5 Sw. J. L. & Trade Am. 323 (1998).

121. C.p.p. art. 498 (Italy).122. Id.


the defense, as in the U.S. adversarial system.123 Civil law countries have alsobegun to eliminate or soften the rule of compulsory prosecution, anothercharacteristic feature of the inquisitorial system; examples include Guate-mala (1994),124 Costa Rica (1998)125 and Chile (2000–2003).126

Important to this study, a number of countries have introduced consen-sual mechanisms inspired by plea bargaining, including Germany (duringthe 1970s), Italy (1989), Guatemala (1994),127 Argentina (1998), Costa Rica(1998),128 and France (1998).

There have also been signiªcant moves toward the coordinate ideal. Anumber of countries have introduced the jury into their criminal process,such as Russia (1993)129 and Spain (1995).130 A substantial number of civillaw jurisdictions have also tried to introduce more oral case-managementtechniques from the coordinate model and to reduce the traditional impor-tance of the written dossier in their inquisitorial systems. Examples of thesejurisdictions include Italy (1989),131 Argentina (federal system, 1992),132

Guatemala (1994),133 the province of Buenos Aires in Argentina (1998),Paraguay (1999),134 Venezuela (1999),135 and Chile (2000–2003).136

Both with respect to the model of the dispute and the coordinate model,recent reforms in civil law jurisdictions seem to conªrm the weak version ofthe Americanization thesis in criminal procedure. However, they do not nec-essarily conªrm the strong version of the Americanization thesis, which positsthat the legal practices and cultures of the inºuenced jurisdictions will even-tually resemble or imitate their American counterparts. A rigorous assess-ment of the strong thesis requires more detailed study of how legal ideas,rules and institutions travel or circulate among legal systems as well as thekinds of transformations that occur during this process. In other words, be-

123. Id. arts. 493, 498.124. Cód. Proc. Pen. art. 25 (Guat.).125. Cód. Proc. Pen. art. 22 (Costa Rica).126. Cód. Proc. Pen. art. 170 (Chile).127. For an analysis of this mechanism in Guatemala, see Alberto Bovino, Temas de Derecho

Procesal Penal Guatemalteco 141–64 (1996).128. Cód. Proc. Pen. arts. 373–75 (Costa Rica).129. See Gary S. Gildin, Trial by Jury in the New Russia: A Travelogue, 15 Dick. J. Int’l L. 151 (1996);

Steven R. Plotkin, The Jury Trial in Russia, 2 Tul. J. Int’l & Comp. L. 1 (1994); Stephen C. Thaman, TheResurrection of Trial by Jury in Russia, 31 Stan. J. Int’l L. 61 (1995). See also Leonard Orland, A RussianLegal Revolution: The 2002 Criminal Procedure Code, 18 Conn. J. Int’l L. 133 (2002) (indicating that thenew Russian Criminal Procedure Code of July 2002 also includes trial by jury for serious cases).

130. See Carmen Gleadow, History of Trial by Jury in the Spanish Legal System 245–85

(2000); Stephen C. Thaman, Spain Returns to Trial by Jury, 21 Hastings Int’l & Comp. L. Rev. 241

(1998). See also Stephen C. Thaman, Europe’s New Jury Systems: The Cases of Spain and Russia, 62 Law &

Contemp. Probs. 233 (1999).

131. See, e.g., Grande, supra note 32, at 237–39, 257.132. See, e.g., Guariglia & Bertoni, supra note 117, at 64–65.133. See Ramírez García & Urbina, supra note 116, at 480–81.134. See, e.g., Alfredo Enrique Kronawetter, Paraguay, in Las Reformas, supra note 116, at 605, 635–

36.135. See, e.g., Brown Cellino, supra note 119, at 806–07.136. See, e.g., Riego, Chile, in Las Reformas, supra note 116, at 167, 179.


sides studying and identifying the inºuences of the Anglo-American system,it is necessary to understand how these inºuences have been translated andwhat kinds of interactions have occurred between these reforms and the pre-existing practices of the receiving criminal justice systems.

V. The Circulation of Legal Ideas: From Legal Transplants to

Legal Translations

Until now, the metaphor of the legal transplant—popularized by AlanWatson—has been the dominant metaphor in analyzing the circulation oflegal institutions among legal systems.137 Through numerous books andarticles, Watson has shown that the transfer, borrowing, and imposition oflegal rules have been common since time immemorial, and he has used themetaphor of the transplant to explain such occurrences.138 The success ofWatson’s proposal is attributable to a number of causes, and multiple com-mentators in very diverse ªelds of law have adopted it.139 Nonetheless, inthis Part, I challenge the metaphor’s adequacy in discussing the circulationof legal institutions and instead propose the model of the legal translation asa new heuristic device for approaching these issues.

There are a number of reasons for the success of the transplant metaphor.First, the substantial increase in the circulation of legal ideas and institu-tions because of globalization,140 has created a need for and interest in waysof conceptualizing this phenomenon; the transplant metaphor’s popularity is

137. See, e.g., Alan Watson, The Evolution of Law (1985); Watson, supra note 22; Alan Wat-

son, Society and Legal Change (1977); Alan Watson, Aspects of Reception of Law, 44 Am. J. Comp. L.

335 (1995).138. The work of Watson should be partially read as an intervention in debates about the autonomy of

law. In his historical analyses, Watson has attempted to demonstrate that law is not a mirror of socialculture, society, or economy, but rather that law changes as a consequence of internal developments inwhich the transplant of legal rules has a central role. In this way, Watson tries to challenge cultural analy-ses a là Montesquieu, Marxism, and contemporary “law & society” approaches. For critical assessments ofWatson’s intervention in these debates, see Richard L. Abel, Law as Lag: Inertia as a Social Theory of Law,80 Mich. L. Rev. 785 (1982); William Ewald, The American Revolution and the Evolution of Law, 42 Am. J.

Comp. L. 1 (1994); William Ewald, Comparative Jurisprudence (II): The Logic of Legal Transplants, 43 Am. J.

Comp. L. 489 (1995).139. See, e.g., Ajani, supra note 22; Berkowitz et al., supra note 22 at 5; Buscaglia & Raºiff, supra

note 22; Ugo Mattei, supra note 22; Frederick Schauer, supra note 22.

140. See, e.g., Lawrence Friedman, Erewhon: The Coming Global Legal Order, 37 Stan. J. Int’l L. 347

(2001). Globalization has also been characterized by the increasing circulation of goods and services aswell as information. See generally Saskia Sassen, Globalization and its Discontents: Essays on

the New Mobility of People and Money (1998). For general introductions to globalization debates,see, for example, The Ethical Dimensions of Global Change (Barry Holden ed., 1996); Global

Culture: Nationalism, Globalization, and Modernity (Mike Featherstone ed., 1990); Global

Democracy: Key Debates (Barry Holden ed., 2000); The Global Transformations Reader: an

Introduction to the Globalization Debate (David Held & Anthony McGrew eds., 2000); The

Globalization Reader (Frank J. Lechner & John Boli eds., 2000); Globalizing Institutions: Case

Studies in Regulation and Innovation (Jane Jenson & Boaventura de Sousa Santos eds., 2000);Governance in a Globalizing World, supra note 22. On the potential effects of globalization ten-dencies on the relationship between state sovereignty and international law, see Kal Raustiala, The Archi-tecture of International Cooperation: Transgovernmental Networks and the Future of International Law, 43 Va. J.

of Int’l L. 1 (2002).


due, in part, to the necessity of ªlling the gap between theory and practice.Second, the transplant metaphor is powerful because of its inherently com-parative nature: the transplant, as a medical or botanical metaphor, supposesan original body or environment and a receiving one. Thus, the metaphorallows for the comparison between both the original and receiving legal sys-tems as well as comparison between the original and transplanted legalrules, ideas or institutions. This can be a particularly interesting approach tocomparative law and a useful way to analyze the phenomena of cultural legalinºuences—like Americanization—and other theses concerning the trend oflegal globalization (such as the convergence thesis).141 Finally, the idea of thetransplant is also powerful because, as a medical and botanical metaphor, itincludes the transferred legal rule’s necessity for adjustment to the new or-ganism or environment—the practices of an existing legal system—and, atthe same time, the possibility of rejection by the receiving organism or envi-ronment—the receiving legal system.142

Nevertheless, the metaphor of the legal transplant presents several short-comings. It fails to account for the possibility that, in many cases, legal con-cepts and practices are transferred on some conceptual levels but not others.For example, constitutional review is an idea and an institutional mechanismthat was “transplanted” from the United States to continental Europe.143 How-ever, the system of constitutional review in continental Europe presents sub-stantial differences from the American version: (a) whereas in the UnitedStates every court can hold certain statutes or state practices to be unconsti-tutional, in continental Europe, generally only one centralized constitutionalcourt can do it; (b) while in the United States the courts that serve the con-stitutional judicial review function are part of the judiciary, in continentalEurope this is not always the case.144

The metaphor of the transplant is not ºexible enough to capture thesesubtleties and conveys the misleading notion that the same concepts andpractices of “constitutional review,” exist in both the United States and con-tinental Europe. A kidney or an elm will look essentially alike in its originaland receiving body or environment, but this frequently does not happen

141. See The Gradual Convergence (B.S. Markesinis ed., 1994); Pierre Legrand, European Legal

Systems Are Not Converging, 45 Int’l & Comp. L.Q. 52 (1996). For discussion of the convergence thesis incriminal procedure, see, for example, Diane Marie Amann, Harmonic Convergence? Constitutional CriminalProcedure in an International Context, 75 Ind. L.J. 809 (2000); Craig M. Bradley, The Convergence of theContinental and the Common Law of Criminal Procedure, 7 Crim. L.F. 471 (1996) (reviewing Criminal

Justice in Europe: A Comparative Study (Phil Fennell et al. eds., 1995)); Nico Jörg et al., Are In-quisitorial and Adversarial Systems Converging?, in Criminal Justice in Europe, supra.

142. Not surprisingly, the most interesting debate on legal transplants during the 1970s was aboutthe potential rejection of transplanted legal rules or ideas. Compare Otto Kahn-Freund, On Uses and Mis-uses of Comparative Law, in Selected Writings 294, 310 (1978), with Alan Watson, Legal Transplantsand Law Reform, 92 Law Q. Rev. 79 (1976).

143. See Blankenburg, supra note 16.144. For an analysis of the French Constitutional Council, see, for example, Doris Marie Provine,

Courts in the Political Process in France, in Courts, Law, and Politics 177, 190–93.


with legal institutions and ideas, which are imitated at certain conceptuallevels but not at others.

Another problem with the metaphor of the transplant is that even whenthe reformers try to imitate a legal idea or practice as closely as possible, thisnew legal idea may still be transformed by the structure(s) of meaning, indi-vidual dispositions, institutional and power arrangements, systems of incen-tives, etc., present within the receiving legal system.145 For instance, as dis-cussed before, Italian reformers tried to import an adversarial trial by intro-ducing the principle that evidence produced at trial has to be requested bythe parties before its production can be ordered by the judge.146 However,the criminal procedure code also provided an exception to this principle,allowing the trial court, if absolutely necessary, to order, even sua sponte, theproduction of new evidence.147 While this exception is not unusual in U.S.rules of procedure,148 it is rarely exercised by American judges.149 Yet in It-aly, because a substantial number of legal actors are accustomed to the in-quisitorial model (and thus possess a predominantly inquisitorial set of in-ternal dispositions) this exception has been interpreted as giving the courtlarge scope to order new evidence.150 The result seems to be a trial that issometimes interpreted through the model of the dispute—as was originallyenvisioned by the reformers—but is as often interpreted and understoodthrough the model of the ofªcial investigation.151

Given the theoretical framework developed in Part II, this transformationis not surprising in a country such as Italy, where the inquisitorial systemwas clearly predominant as a structure of interpretation of meaning, in theindividual dispositions of most legal actors, and in the distribution of proce-dural powers. However, the point I want to emphasize here is that the meta-phor of the transplant seems to be, again, too rigid to capture the transfor-mations to which the U.S. trial has been subjected in Italy as a consequence

145. This is why Pierre Legrand has pointed out that legal transplants are impossible. The meaning of

a legal rule is not determined—or not only determined—by the words that express that rule, but by therest of the context of meaning—i.e., the legal system—in which this rule is situated. Therefore, everytime that words which express a legal rule are transferred from legal system A to legal system B, themeaning of these words will change because there are no two legal systems exactly alike in regard tocontexts of meaning. The transformation of the transferred rule—and thus the impossibility of the legaltransplant—will be even greater when moving from a common to a civil law system or vice versa, giventhe traditionally deep differences between both kinds of legal systems. See Pierre Legrand, The Impossibilityof ‘Legal Transplants,’ 4 Maastricht J. Eur. & Comp. L. 111 (1997).

146. C.p.p. arts. 190.1, 493.1 & 495.1. (2002) (Italy).147. C.p.p. art. 507. (2002) (Italy).148. See, e.g., Fed. R. Evid. 614(a).149. See United States v. Ostrer, 422 F. Supp. 93 (S.D.N.Y. 1976) (stating that although the court has

discretionary power to call a court witness, this power is rarely invoked).150. See, e.g., Corte costituzionale [Corte cost.] [Constitutional Court], 24 mar. 1993, n.111, Raccolta

ufªciale delle sentenze e ordinanze delle Corte costituzionale [Racc. uff. corte cost.] 1993, vol. 106, 733(Azzari); Cass. Pen., sez. un., 28 dic. 1999, n.5549, (Paternò); Cass. Pen., sez. un. 6 nov. 1992, n.11227,(P.M. in proc. Martin), in Cassazione Penale 1993, vol. 33, 280, n.157; Grande, supra note 32, at 246(citing additional decisions).

151. See Grande, supra note 32, at 246–47.


of its interactions with pre-existing, predominantly inquisitorial, criminaljustice practices.

A third problem with the metaphor of the transplant is that the transfer-ence of legal rules, ideas and practices, may produce a deep transformationnot only in the transferred practice itself but also in the receiving legal sys-tem as a whole. For instance, in the previous example, not only has the im-ported U.S. trial been transformed in its new Italian context, but it has alsointroduced some elements of the dispute into the Italian system. If theseelements are incorporated into the internal dispositions of judges, prosecu-tors, and defense attorneys, this may produce changes in a system where theinquisitorial structure of meaning has been clearly predominant; this may, inturn, redeªne the way power is distributed among legal actors. The meta-phor of the transplant does not seem to be ºexible enough to capture thisphenomenon either.152 Even if a human body has to adjust itself to a neworgan, it will still remain essentially unchanged. The changes produced in alegal system by the transference of legal rules, ideas, and institutions, how-ever, may go much deeper than that.

As a consequence of the legal transplant metaphor’s limitations, GüntherTeubner has proposed the expression “legal irritant” to capture this phe-nomenon.153 The legal irritant metaphor clearly avoids most of the problemsassociated with the transplant metaphor. Particularly, it is able to convey theidea that the transfer of the legal rule or doctrine may prompt a series oftransformations in the receiving legal and social systems. Nevertheless, thelegal irritant metaphor presents important shortcomings, the foremost ofwhich is that it loses the comparative dimension that has made the meta-phor of the transplant so powerful. An irritant does not necessarily comefrom another (legal) system or from outside the system it irritates. Thus, thecomparative dimension of the metaphor is lost regarding both between theoriginal and receiving legal systems and between the original idea or prac-tice and the transferred one.

Given the limitations of both the transplant and irritant metaphors, Ipropose the metaphor of the translation as a superior heuristic device toanalyze the circulation of legal ideas, rules, practices, and institutions.154

152. Understood in botanic terms, the metaphor of the transplant may be more able to answer this

particular critique. The transplantation of a certain species from environment A to environment B canproduce deep changes in the latter.

153. Teubner, supra note 45.154. The metaphor of translation has not been used in approaching the analysis of the circulation of

legal ideas, rules, practices, and institutions between legal systems, though different commentators haveused the metaphor for analyzing other legal phenomena. See, e.g., James Boyd White, Justice as

Translation: An Essay in Cultural and Legal Criticism (1990); Gerald Torres, Translation andStories, 115 Harv. L. Rev. 1362 (2002). Lawrence Lessig has used the metaphor to propose a new theoryof constitutional interpretation. See Lawrence Lessig, Fidelity and Constraint, 65 Fordham L. Rev. 1365

(1997); Lawrence Lessig, Fidelity in Translation, 71 Tex. L. Rev. 1165 (1993); Lawrence Lessig, The Puz-zling Persistence of Bellbottom Theory: What a Constitutional Theory Should Be, 85 Geo. L.J. 1837 (1997);

Lawrence Lessig, Translating Federalism: United States v. Lopez, 1995 Sup. Ct. Rev. 125 (1995); LawrenceLessig, Understanding Changed Readings: Fidelity and Theory, 47 Stan. L. Rev. 395 (1995); For critical


The metaphor of the translation retains the comparative dimension that hasmade the metaphor of the transplant so powerful and that the legal irritantmetaphor lacks. With respect to legal systems, the translation metaphor distin-guishes the source language or legal system—where the legal idea or institu-tion comes from—from the target one—into which the legal idea or institu-tion is translated.155 The translation metaphor also allows a distinction to bemade between the original “text”—the legal idea or institution as developedin the source legal system—and the translated text.

In drawing our attention to the differences between the original andtranslated text, the metaphor of the translation also distinguishes the trans-formations the legal idea may undergo when initially transferred from thesource to the target legal system.156 The ªrst set of transformations ªnds itsorigin in the skills and decisions of the translators,157 i.e., legal reformers.158

The theory and history of translation have presented three main approachesto translation: (1) strict literalism, a “word-by-word matching” between theoriginal and the translated texts; (2) “faithful but autonomous restatement,”where the translator still tries to be faithful to the original but composes, atthe same time, a text that is equally powerful in the target language; and(3) substantial recreation, variations, etc., where the idea of ªdelity to theoriginal is weakened or directly disappears, and the focus is to create a textthat is powerful or appealing in the target language.159 Taking the examplesexamined in this Part, the translation from the United States to Italy of theprinciple that evidence is produced at trial at the request of the parties, in-stead of by order of the judge sua sponte, falls between the ªrst and the sec-

assessments of Lessig’s proposal, see, for example, Steven G. Calabresi, The Tradition of the Written Consti-tution: A Comment on Professor Lessig’s Theory of Translation, 65 Fordham L. Rev. 1435 (1997); Michael C.Dorf, Recipe for Trouble: Some Thoughts on Meaning, Translation and Normative Theory, 65 Fordham L. Rev.

1857 (1997); Sanford Levinson, Translation: Who Needs It?, 85 Geo. L.J. 1457 (1997).

155. Pierre Bourdieu makes a distinction between ªeld of production and ªeld of reception in order tocapture the circulation of ideas between countries in areas such as philosophy, social sciences and litera-ture. See Pierre Bourdieu, The Social Conditions of the International Circulation of Ideas, in Bourdieu: A

Critical Reader 221 (Richard Shusterman ed., 1999). The distinction between production and recep-tion has been widely used in legal analyses. See, e.g., La Réception des Systèmes Juridiques: Implan-

tation et Destin (Michel Doucet & Jacques Vanderlinden eds., 1994); Diego López-Medina, Compara-tive Jurisprudence: Reception and Misreading of Transnational Legal Theory in Latin America 7–12(2001) (unpublished S.J.D. dissertation, Harvard University) (on ªle with Harvard Law School Library);Maier, supra note 63, § 5(D)(5), (D)(8), at 303, 329 and § 8(A), at 811.

156. I refer here to the transformations of the legal idea or institution when initially transferred to thetarget legal system. As we will see, there are transformations that happen once the legal idea or institu-tion has been incorporated into the target legal system.

157. The skills of the translator may include not only his ability to translate (legal) texts, but also hisknowledge of the source (legal) system. For a classic reference on this issue, see John Dryden, On Transla-tion, in Theories of Translation 17, 30 (Rainer Schulte & John Biguenet eds., 1992).

158. The analysis of the work of translators and their decisions and motivations is useful to explainnot only the transformations that the original text may undergo, but also why the translation—i.e., legalreform—happened in the ªrst place. There has been a recent tendency in translation studies to fore-ground the motivations and powers of translators. See, e.g., Susan Bassnett, Translation Studies 6,

10 (3d ed. 2002); Lawrence Venuti, The Translator’s Invisibility (1995).

159. See George Steiner, After Babel: Aspects of Language and Translation 266 (3d ed.1998).


ond approaches.160 The translation of the American practice of constitutionalreview to the continental European context is closer to the third.

The second set of initial text transformations may have its origin in dif-ferences between the source and target languages—the source and the targetstructures of interpretation and meaning. For instance, the word “derecho” inSpanish means both “right” and “law” in English. Therefore, if we want totranslate it from Spanish to English, something will probably be lost alongthe way in choosing only one of the word’s two meanings.161

Moreover, the metaphor of the translation also captures the transforma-tions that the legal idea or practice may undergo in its exchanges with thetarget legal system after its initial translation. These transformations mayinclude the total neutralization of the translated “text”—the legal rule orpractice—by either ostracism (disuse or desuetude)162 or censorship (i.e.,stating that the practice is unconstitutional).163 In addition, these transfor-mations may also include a struggle between different actors and groupswithin the target legal system over the meaning of the translated institu-tion.164

Finally, the metaphor of the translation is also apt to describe the trans-formation that the receiving linguistic and social practices may undergo un-

160. Id. at 266 (“The dividing lines between the three types are necessarily blurred.”). However, the

distinction between the three types can still be useful to analyze processes of translation.161. For analyses of the obstacles that the differences between languages present for translations, see,

for example, Roman Jakobson, On Linguistic Aspects of Translation, in Theories of Translation 144,

146–51 (Rainer Schulte & John Biguenet eds., 1992). The idea of untranslatability of texts ªnds itsorigin in these obstacles because it assumes that these differences between languages make translationsimpossible. However, most translation studies today reject this idea of untranslatibility. See, e.g., HugoFriedrich, On the Art of Translation, in Theories of Translation, supra, at 11, 14–15.

162. An example of this case of neutralization is the German StPO section 239, which states inter-rogatories at trial shall be developed through direct and cross-examination at the request of the prosecu-tion and the defense, a rule translated from the Anglo-American adversarial system. However, this rule israrely applied since, according to Roxin, it does not ªt within the structure of German criminal proce-dure because it takes away the managing of the trial from the presiding judge. See Roxin, supra note 66,§ 42, at 343.

163. An example of this is the 1989 “translation” of American rules of evidence into Italian criminalprocedure, like hearsay, which allowed the in-trial use of depositions collected during the pre-trial phaseonly for impeachment purposes. The Italian Constitutional Court, in its decision 255/1992, held uncon-stitutional art. 500.3 that had introduced such a limitation. Corte cost., 18 mag. 1992, n.255, 104 Racc.uff. corte cost. 1992, 7 [hereinafter Decision 255/1992]. For a more detailed analysis of this issue, seeinfra note 223.

164. For instance, the American practice of the defense team making its own pre-trial investigationwas translated to the Italian practices in 1989. D.L.vo 28 lug. 1989, n.271, art. 38, Raccolta Ufªcialedegli Atti Normativi della Repubblica Italiana [Racc. Uff.] 1989, vol. 8, 3772, 3780, published inGazzetta Ufªciale della Repubblica Italiana [Gazz. Uff.] 5 ago. 1989, n.182 [hereinafter Law 271/1989].Since these practices went against the inquisitorial structure of meaning and internal dispositions of asubstantial number of legal actors, these actors interpreted that the elements of proof collected by thedefense attorney had less probative value than the ones collected by the prosecutor. Conversely, the sup-porters of the adversarial reform maintained that the elements of proof collected by the defense had thesame probative value as those collected by the prosecutor. The latter ªnally imposed their interpretationof the translated practice with the enactment of articles 391-bis-decies of the C.p.p. in D.L.vo 7 dic. 2000,n.397, published in Gazz. Uff. 3 gen. 2001, n.2 [hereinafter Law 397/2000].


der the inºuence of the translated text.165 Consider the inºuence that thetranslation of texts by such thinkers as Beccaria, Montesquieu, and Voltaireinto English had, not only on political vocabulary, but also on the politicalthought of Revolutionary America.166 Similarly, the translation into Frenchof Blackstone’s Commentaries on the Laws of England (1765) had an impact notonly on the French legal vocabulary but also resulted in the positive estima-tion of the English system by a number of French actors, which culminatedin the importation of the trial by jury two years after the French Revolu-tion.167 The translation of legal ideas and institutions between legal systemscan also have an impact not only on the vocabulary but also on the real prac-tices of the receiving system—i.e., moving formerly inquisitorial practicesin the direction of the adversarial system.

VI. Plea Bargaining as a Trojan Horse of the Adversarial System

American plea bargaining is a procedural mechanism through which theprosecution and defense can reach an agreement for the disposition of a case,subject to the approval of the court.168 The agreement may present itself inseveral forms, but it usually consists of the defendant pleading guilty to anoffense or a number of offenses. In exchange, the prosecutor drops othercharges, accepts that the defendant pleads guilty to a lesser offense, or re-quests that—or does not object if—the defendant receives a certain sen-tence.169

The importation of plea bargaining exempliªes the translation of a legalpractice that could potentially Americanize inquisitorial jurisdictions. Thereare few mechanisms or institutions more characteristic of the U.S. adversarialsystem and, more speciªcally, of the model of the dispute.170 By deªnition,

165. A number of recent translation studies have focused on this issue and have showed the transfor-

mations that translated texts may cause in the receiving linguistic and social practices. See, e.g., LinKenan, Translation as a Catalyst for Social Change in China, in Translation and Power 160 (MariaTymoczko & Edwin Gentzler eds., 2002); Alexandra Lianeri, Translation and the Establishment of LiberalDemocracy in Nineteenth-Century England: Constructing the Political as an Interpretive Act, in Translation

and Power, supra, at 1.166. On the inºuence that these thinkers had on Revolutionary America, see Bernard Bailyn, The

Ideological Origins of the American Revolution 26–30 (enlarged ed. 1992).167. On the translation of Blackstone’s Commentaries to French and their inºuence on the positive per-

ception of the English system by a number of French actors, see Antonio Padoa Schioppa, I Philosophes e laGiuria Penale, I–II Nuova rivista storica 107, 123–28 (1986).

168. See, e.g., Fed. R. Crim. P. 11(e).169. See, e.g., Fed. R. Crim. P. 11(e)(1).170. Recall that in this Article, I use the term “adversarial” in a descriptive way. At the descriptive

level, since plea bargains are widely practiced in the U.S. system, it is reasonable to consider this practicehighly characteristic of U.S. criminal procedures. In addition, this practice also matches well with theconception of criminal procedure as a dispute between two parties before a passive umpire, a conceptionthat extends to the American criminal justice system. If instead of using this expression in a descriptiveway, one understands the adversarial system as a prescriptive ideal—according to which, for instance, allor most cases should be decided through a trial by jury—one could consider that plea bargains are notpart of the American adversarial system, but rather, are undermining it. As mentioned before, the criti-cisms in the United States against plea bargaining have been very important among commentators and


American plea bargaining, assumes an adversarial conception of criminalprocedure as a dispute between two parties facing a passive decision-maker.It makes sense in a dispute model that the parties be allowed to reach anagreement over a plea bargain. That is, the parties may negotiate in order toreach such an agreement, and if the parties agree that the dispute is over, thedecision-maker should not have any power (or only a relatively minor andformal power) to reject this decision.171

Furthermore, features from the model of the dispute are conditions thatenable the possibility of developing plea bargaining. For example, the guiltyplea, which enables the defendant to end the determination of guilt or inno-cence, has provided the defense with a bargaining tool in its negotiationswith the prosecution. In addition, the fact that the judge, as a passive deci-sion-maker, usually accepts the agreement reached by the parties (the realowners of the process) also provides an incentive for the development of suchpractices.172 Given the usual deference of the judge toward the requests ofthe parties, the defendant can be relatively certain that the bargain struckwith the prosecutor will be fulªlled, even in those situations in which thejudge may not be limited by the requests of the parties, as with certain sen-tencing bargains.173 In addition, because the prosecutor has the power todrop charges or lessen a charge, the broad prosecutorial discretion found inthe model of the dispute provides the prosecutor with powerful and ºexibletools to negotiate with the defense toward a guilty plea.174

Conversely, there are few practices that are more incompatible with theinquisitorial system and the model of the ofªcial investigation than plea

have been based, at least to a certain extent, on this prescriptive conception of the adversarial system. See,e.g., Alschuler, supra note 111; Alschuler, supra note 55; John H. Langbein, Torture and Plea Bargaining,46 U. Chi. L. Rev. 3; Stephen J. Schulhofer, Criminal Justice Discretion as a Regulatory System, 17 J. Legal

Stud. 43; Stephen J. Schulhofer, Is Plea Bargaining Inevitable?, 97 Harv. L. Rev. 1037; Stephen J. Schul-hofer, Plea Bargaining as Disaster, 101 Yale L.J. 1979. For defenses of plea bargaining in the UnitedStates, see, for example, Frank H. Easterbrook, Criminal Procedure as a Market System, 12 J. Legal Stud.

289; Frank H. Easterbrook, Plea Bargaining as Compromise, 101 Yale L.J. 1969; Robert E. Scott & Wil-liam J. Stuntz, Plea Bargaining as Contract, 101 Yale L.J. 1909. It is important to emphasize, then, thatthese criticisms of plea bargaining and my statement that plea barganining is characteristic of the Ameri-can adversarial system are not incompatible, because while in most of these criticisms, the adversarysystem is considered a prescriptive ideal, I use it in this work just as a descriptive category. In otherwords, one could state that plea bargaining is a very distinctive characteristic feature of the U.S. adver-sarial system from a descriptive perspective and still think that plea bargaining is problematic from anormative or prescriptive perspective.

171. In the U.S. federal system, the court must advise the defendant about the waiver of rights, checkthe voluntariness of the guilty plea, and determine whether there is a factual basis for it. See generally Fed.

R. Crim. P. 11. However, courts usually defer to agreements reached by the parties.172. Conversely, an extended practice of plea bargaining between the parties also reinforces the con-

ception of the judge as a passive umpire.173. See, e.g., Fed. R. Crim. P. 11(c), that includes two kinds of sentencing recommendations that the

prosecutor can make: the ªrst does not bind the court at all while the second binds the court only after ithas accepted the plea agreement. For a proposal to allow the prosecutor to set the maximum sentence, seeScott & Stuntz, supra note 170, at 1953–57.

174. Commentators have identiªed this link between prosecutorial discretion and plea bargains. See,e.g., Tracey L. Meares, Rewards for Good Behavior: Inºuencing Prosecutorial Discretion and Conduct with Finan-cial Incentives, 64 Fordham L. Rev. 851, 862–73 (1995).


bargaining. First, the very concept of the guilty plea does not exist in theinquisitorial system.175 Second, there are no two parties who negotiate andbargain, as in the adversarial system, and who could reach a compromise notonly about their respective claims but also about the facts of the case. In themodel of the ofªcial investigation, the prosecutor is not a party to the casebut rather another ofªcial, who, like the judge, has to determine what hashappened.176 In this model, the “real” truth has to be determined by theprosecutor; it cannot be negotiated and compromised. In any case, the judgehas the ªnal word on the investigation’s conclusions. Furthermore, the veryact of negotiating with the defendant has traditionally been considered im-proper conduct by these ofªcials. In negotiations and bargains, the partieshave to recognize each other as equals, at least on a certain level. But in themodel of the ofªcial investigation, the prosecutor, the judge, and the defen-dant are not equals because the latter has an interest at stake in the processthat the former do not.

Moreover, most of the conditions required for the potential developmentof plea bargaining do not exist in the inquisitorial system. Not only does theconcept of the guilty plea not exist in that system, but the prosecutor alsohas more limited discretion to decide what cases and charges she wants tomove forward, as symbolized and regulated by the rule of compulsory prose-cution.177 Furthermore, judges do not usually feel constrained by requeststhe parties may make regarding sentencing.

Nevertheless, despite this apparent incompatibility between plea bar-gaining and the model of the ofªcial investigation, a substantial number ofcivil law countries have recently shown an interest in translating this mecha-nism into their procedures. The reasons vary from jurisdiction to jurisdic-tion, but one common reason has been increasing crime rates in most ofthese countries in recent years.178 This situation has produced an increasingburden on their criminal procedures, requiring them to handle more crimi-nal cases in less time than before.179 Obtaining a defendant’s consent, there-fore, through negotiations or the offering of beneªts, could render unneces-sary, or provide a justiªcation to simplify or directly avoid, the regular in-quisitorial criminal proceedings. Understood within this context, the intro-

175. See Langbein, supra note 39, at 73–74; Moskovitz, supra note 38, at 1153.176. See Weigend, supra note 35, at 1233–34.177. See supra note 82.178. For an analysis of other reasons that continental European jurisdictions have translated plea bar-

gains to accommodate their procedures, see, for example, Françoise Tulkens, Negotiated Justice, in Euro-

pean Criminal Procedures, supra note 49, at 645–49179. Regarding Argentina, see infra note 284. In France, the project that introduced French “plea

bargaining” was aimed at strengthening the efªciency of the criminal procedure. See Jean Cedras, L’hypothèse del’américanisation du droit pénal français, 45 Archives de Philosophie du Droit 149, 156 (2001). Re-garding Germany, see infra note 184 and accompanying text. Concerning Italy, see, for example, Paolo

Ferrua, La Giustizia Negoziata nella Crisi della Funzione Cognitiva del Processo Penale, 3 Studi sul Proc-

esso Penale 131, 134 (1997).


duction of consensual negotiating mechanisms has been seen as a way ofmaking the rigid inquisitorial systems more ºexible.

The introduction of plea bargaining in continental Europe and LatinAmerica is particularly relevant to an analysis of the Americanization thesis.The very fact that U.S. plea bargaining has become such an important pointof reference in recent reforms in the civil law world supports the weak ver-sion of this thesis. One could also think, though, that the spread of plea bar-gaining would also support the thesis’s strong version. United States pleabargaining assumes an adversarial conception of the criminal process. If civillaw countries translated this mechanism faithfully and did not reject it af-terwards, the effect of this import could be that the prosecution and the de-fense would begin to think of themselves as parties in a dispute and be per-ceived by the other legal actors as such, and something similar would de-velop with the judge as a passive umpire. If this happened, it would alter theinternal dispositions of legal actors and would produce a change or at least aserious challenge to the substructure of meaning of the ofªcial investigationby the substructure of the dispute. This could produce, then, a change in theclassic relations of power from one structure to the other, as well as changesin the organization of human and material resources, including case-manage-ment techniques, among other elements.

In other words, one could think about plea bargaining as a Trojan horsethat can potentially bring, concealed within it, the logic of the adversarialsystem to the inquisitorial one. If this happened, then the Americanizationthesis would be valid even in its strong version because inquisitorial systemsthat translate plea bargaining would gradually become “Americanized” byadopting an adversarial conception of criminal procedure.

The following Parts analyze four speciªc plea bargaining translations thatdemonstrate the complex effects of plea bargaining’s circulation among civillaw jurisdictions. These Parts support the notion that Americanization isnot taking place in all the jurisdictions that have developed mechanismsinspired by plea bargaining in recent years, even if it may be happening insome of them. The four examples analyzed are the so-called Absprachen inGermany, the patteggiamento in Italy, the procedimiento abreviado in Argentina,and the composition in France. These four mechanisms either have been ex-plicitly labeled as “plea bargaining” or have been considered close to it ineach of these four countries and, in some cases, in the United States.180

Analysis of each of these translations is divided into two main parts. Theªrst part analyzes the similarities and differences between U.S. plea bar-gaining and the Absprachen, patteggiamento, procedimiento abreviado, and compo-

180. Regarding the Absprachen, see, for example, Herrmann, Bargaining Justice, infra note 183;

Schünemann, Die Absprachen im Strafverfahren, infra note 186, at 527. Regarding the pattegiamento, see, forexample, Ferrajoli, supra note 56. Regarding the procedimiento abreviado, see, for example, Gabriel IgnacioAnitua, El juicio abreviado como una de las reformas penales de inspiración estadounidense que posibilitan la expan-sión punitiva, in El Procedimiento Abreviado, supra note 23, at 137, 144 (Julio B. J. Maier & AlbertoBovino eds., 2001). Regarding the composition, see, for example, Merle & Vitu, supra note 109, at 396.


sition. The second part analyzes the kinds of transformations these transla-tions may produce in the German, Italian, Argentine, and French criminalprocedures. The central question in this part is whether these inquisitorialprocedures are keeping their traditional model of the investigation, whetherthey are moving in the direction of the American model of the dispute, orwhether they are moving in a direction different from either one of thesetwo. This study will enable us to examine what changes these reforms maybe producing in the criminal procedures of the civil law tradition.

VII. German “Plea Bargaining”

As one of the jurisdictions where the model of the ofªcial investigationprevailed, the German system did not lend itself to the possibility of bar-gaining with the defendant.181 In fact, at the end of the 1970s, one of themost distinguished American comparativists referred to Germany as a “landwithout plea bargaining.”182 Nevertheless, during the 1970s, judges, prose-cutors and defense attorneys of the German criminal justice system begandeveloping bargains, or Absprachen, on the quiet before and during trial.183

Inºuencing this development was the increasing number of criminal casesand the difªculties and length of criminal trials for economic, environ-mental, and drug offenses.184 In this sense, the practice seems to have arisenas a response to practical needs, rather than as a product of deep culturalinºuences of the American system over the German one. The practice waskept quiet until 1982, when the ªrst article on the subject was publishedunder a pseudonym, indicating that this was a delicate practice to discuss,given that it went against the basic assumptions of the German criminalprocedure.185 Since then, a lively debate has ensued, in which importantcommentators have criticized and opposed the practice while courts havegenerally upheld it, though imposing certain limits to the practice.186

181. See, e.g., Hans-Heiner Kühne, Germany, in Criminal procedure systems in the European

Community 137, 145 (Christine Van Den Wyngaert et al. eds., 1993) (“The principle of ‘instruction’(Instruktionsmaxime), as opposed to the principle of negotiation (Verhandlungsmaxime), imposes upon allauthorities involved in criminal proceedings the duty to search for the truth.”). The StPO establishes theprinciple of “instruction” or inquisitorial principle. §§ 155 Nr. 2, 244 Nr. 2 StPO (F.R.G.).

182. Langbein, supra note 54.183. See Joachim Herrmann, Bargaining Justice—A Bargain for German Criminal Justice?, 53 U. Pitt.

L. Rev. 755, 755–57 (1992).

184. See Roxin, supra note 66, § 15, at 95; Frase & Weigend, supra note 111, at 345 n.195.185. Detlef Deal, Der strafprozessuale Vergleich, 2 Strafverteidiger 545 (1982).186. For an analysis of this debate in English, see Thomas Swenson, The German “Plea Bargaining” De-

bate, 7 Pace Int’l L. Rev. 373 (1995). One of the most important critics of these practices has been Prof.Bernd Schünemann. See, e.g., Bernd Schünemann, Absprachen im Strafverfahren? Grundlagen, Gegenständeund Grenzen, in 2 Verhandlungen des Achtundfünfzigsten Deutschen Juristentages L6(Ständige Deputation des deutschen Juristentages ed., 1990). The agreements have not been formallyintroduced in the German Criminal Procedure Code. See, e.g., Thomas Weigend, Abgesprochene Gerechtigkeit—Efªzienz durch Kooperation im Strafverfarhern?, 45 JuristenZeitung [JZ] 774, 775 (1990) (stating thatthe German Criminal Procedure Code does not mention the agreements).


Although the form and content of German negotiations and agreementsvary, the basic idea is the following: during trial preparation or during trial,the defendant may make an offer to confess at trial, in exchange for a guar-antee by the judge that the sentence will not exceed a certain limit or that anumber of charges will be dismissed by the prosecutor.187 These bargainsmay be initiated by the defense, the judge, or the prosecutor. Not all of theactors must participate in the bargaining, so they may develop between thetrial judge and the defense.188 However, all of the actors in the case must beinformed of any bargains in open court.189 Furthermore, the agreements re-quire cooperation and mutual trust between the judge, the prosecution andthe defense.190

As is clear from this description, the transformations that U.S. plea bar-gaining underwent when translated into German criminal procedure aresubstantial.191 First, German bargains or Absprachen are not about guiltypleas but about confessions. Therefore, by making it easier to prove the in-dicted facts through a confession, the agreement generally shortens the trialbut does not replace the trial altogether.192 Second, since the defense has fullaccess to the written dossier that contains the pre-trial investigation—thereis full discovery from the prosecution to the defense in the German sys-tem—the defense has better knowledge of the case than its U.S. counterpartduring the negotiations.193 Third, the presiding trial judge is usually an ac-

187. In this Article, I will only analyze the types of agreements that have generated the most contro-

versy within the German criminal justice system. There are two other mechanisms that have also beencompared to American plea bargaining. The ªrst is the penal order procedure—Strafbefehlsverfahren—regulated in sections 407–12 of the StPO. In this procedure the prosecutor, instead of moving the casefor a trial, applies to the judge for a penal order. The prosecutor prepares a draft of the penal order whichdescribes the details of the case and which requests a speciªc legal consequence that may include a ªne, awithdrawal of a permission to drive, or even imprisonment up to one year to be served on probation. Ifthe draft penal order is validated by the judge, it is sent to the defendant who has to oppose it in order toavoid its execution. This procedure is used to handle the bulk of misdemeanor cases. For a description ofthis in English, see Herrmann, supra note 183, at 760–63. This procedure has been compared to pleabargaining because the prosecutor and the defense are allowed to negotiate regarding the consequence ofthe penal order. See, e.g., Frase & Weigand, supra note 111, at 345. The second procedure that has beencompared to the plea bargain is the one established in § 153(a) of the StPO. In a case involving a lessserious criminal offense, with the consent of the court and the consent of the accused, this procedureauthorizes the public prosecution ofªce to dispense with preferment of public charges and concurrentlyimposes a condition upon the accused. This condition may include repairing the damage occasioned bythe offense, performing a non-proªt service, etc. In this sense, it is similar to the U.S. practice of diver-sion. This procedure was added to the German Criminal Procedure Code in 1975, see Herrmann, supranote 183, at 757. For an analysis of this mechanism in English, see id.

188. See, e.g., Swenson, supra note 186, at 378–79.189. Entscheidungen des Bundesgerichtshofes in Strafsachen [BGHSt] [Supreme Criminal Court] 43,

195 (195, 205–06) (F.R.G.).190. See, e.g., Swenson, supra note 186, at 378.191. Anglo-American plea bargaining has been a model of reference and a factor that fostered the de-

velopment of these bargaining practices. See Weigend, supra note 186.192. See Herrmann, supra note 183, at 764.193. Id. The U.S. Supreme Court has held that the prosecutor does not have a constitutional duty to

disclose impeachment or afªrmative-defenses evidence before a guilty plea is entered. See United States v.Ruiz, 536 U.S. 622 (2002). For an analysis of the problems that the limited discovery presents to pleabargains in the United States, see, for example, Eleanor J. Ostrow, The Case for Preplea Disclosure, 90 Yale


tive participant in these bargains and agreements. The two main actors ne-gotiating in the German system are not usually the prosecutor and the de-fense as in the U.S. model, but rather the presiding trial judge and the de-fense.194

These substantial differences between U.S. plea bargaining and the Ger-man Absprachen reveal the inadequacy of the “legal transplant metaphor” incapturing this phenomenon of legal inºuences. It is misleading to state thatU.S. plea bargaining has been transplanted to Germany because this state-ment conveys the false notion that the same legal mechanism exists in bothcountries. Conversely, the translation metaphor accounts for the possibilitythat substantial differences can develop between the original and the trans-lated legal practice. Through the metaphor of translation, the differencesbetween U.S. and German “plea bargains” can be explained by the transfor-mations that this mechanism underwent when translated from a system witha prevailing adversarial structure of interpretation and meaning to a systemwhere an inquisitorial structure of meaning dominates. The German “pleabargaining” translators quietly introduced the practice, aware that its le-gitimacy and meaning within the German inquisitorial structure of inter-pretation were questionable. Thus, for example, they had no power to for-mally change other aspects of German criminal procedure through a reformof the German Criminal Procedure Code. These restrictions limited theirrange of decision-making power and precluded the possibility of ªdelity toU.S. plea bargaining model when translating the new practice.

This limited power meant that the translators had to negotiate substan-tially with the pre-existing inquisitorial structure of interpretation andmeaning, predominant internal dispositions among German legal actors,and structure of procedural power. For instance, German criminal proceduredid not include the concept of the guilty plea and there was no statutoryreform that could introduce the concept. Therefore, the translators had totreat the bargained admission of guilt not as a guilty plea but as a confes-sion. This means that trials still have to be conducted, even if they can beshortened through the Absprachen. Moreover, in the inquisitorial structure ofinterpretation and meaning, the judge is an active, and the most powerful,player in the trial.195 Consequently, any workable bargains to dispose of casesin an expedient fashion must include him, and in practice, the judge plays acentral role in negotiations; only the judge can assure the defendant that thesentence will not rise above a certain limit, and only the judge has controlover the power of the prosecutor to dismiss charges in the trial phase. Thus,the translation of plea bargaining to German practices had to be adapted tothe pre-existing distribution of procedural powers where the judge stands asthe most important ªgure at trial.

L.J. 1581 (1981).

194. See Herrmann, supra note 183, at 764.195. See, e.g., Roxin, supra note 66, § 15 at 95.


The next issue to explore is the types of transformations these practicesmay produce in the German criminal procedure in the future. United Statesplea bargaining has the potential to transform Germany’s structure of inter-pretation and meaning, the internal dispositions of its legal actors, and thedistribution of procedural powers among them. In this case, however, theGerman Absprachen does not appear to be a Trojan horse for the U.S. modelof the dispute. The German judge is still the most active player at trial andhas not become a passive umpire. Therefore, even if the widespread adoptionof this practice affects the way German judges perceive their roles and areperceived by other legal actors, the ªnal result will still not be a passive de-cision-maker.196

However, would the judge still be an ofªcial investigator of the truth ifthis practice imposed its logic on the rest of the German criminal justicesystem? In other words, can the model of the ofªcial investigation absorband survive this translation, or will it be radically transformed by it? Eitherscenario is possible.

The model of the ofªcial investigation can absorb confession agreements,so long as most legal actors understand and use the agreements as a tool todiscover the truth instead of as a mechanism to dispose of criminal casesquickly.197 In fact, the offer of beneªts to the defendant in exchange for adetailed confession, which would serve as an attenuating circumstance forpurposes of sentencing, is not new at all within the model of the ofªcial in-vestigation. Thus, confession agreements could be interpreted as an exten-sion of this practice.

Two important court decisions provide indications that a substantialnumber of legal actors in the German criminal justice system are trying tointerpret and drive the confession agreements in this direction. First, theFederal Constitutional Court (BverfG), in its only decision on this issue,198

stated that these agreements are admissible so long as they do not violate the

196. While there are no ofªcial data on how widely this practice is employed, according to Weigend,

various estimates and surveys indicate that about every fourth trial is settled. Weigend, supra note 191, at774. This indicates that judges, prosecutors, and defense attorneys have widely accepted the new practice.The incentives that the new practice generates for judges, prosecutors, and defense attorneys—i.e., thereduction of their caseload and the acceleration of the disposition of cases—seems to be the main reasonfor this widespread acceptance. For an analysis of these incentives, see Bernd Schünemann, Die Absprachenim Strafverfahren, in Festschrift für Peter Riess 525, 533–34 (Ernst-Walter Hanack et al. eds., 2002).

197. The system of the ofªcial investigation could also survive if the bargaining practices were cen-sored—i.e., through a legal prohibition of them. Even if some law professors have suggested this possi-bility, the discussion between judges, prosecutors, and defense attorneys has focused not on the elimina-tion of the bargaining practices, but in how the translated practices should be interpreted. The explana-tion for this limitation of the discussion seems to be rooted in the incentives that judges, prosecutors, anddefense attorneys have had to support the bargaining practices. Arguably because law professors are notunder the same system of incentives, they have been more faithful to their pre-existing inquisitorialinternal dispositions than the practitioners. On the deep division between the positions of most lawprofessors and practitioners regarding the agreements, see, for example, Thomas Weigend, Eine Prozes-sordnung für abgesprochene Urteile?, 19 Neue Zeitschrift für Strafrecht [NStZ] 57 (1999).

198. Bundesverfassungsgericht [BverfGE] [Federal Constitutional Court] (Judgment of Jan. 27,1987), quoted in Volker Gallandi, Absprachen im Strafprozess, 1987 NStZ 419, 419.


constitutional notion of the Rechtsstaat—rule of law—and a number of sub-stantive and procedural principles.199 Among these principles, the BVerfGmentioned the principle of the duty of the judge and prosecutor to investi-gate the material truth.200 Therefore, because of this duty to investigate thetruth, the judge cannot convict a defendant based on a confession agreementif the judge should have considered himself obliged to hear further evidence.201

In addition, the German Federal Supreme Court in its decision of August28, 1997, analyzed in detail for the ªrst time the requirements these agree-ments must meet in order to be admissible.202 Among these requirements,there must be no violation of the judicial duty to determine the truth.203

This means, for instance, that the court cannot base its verdict only on aconfession obtained through a bargain; the court has to examine the credi-bility of such a confession and, eventually, order the production of additionalevidence at trial.204

It is reasonable to say, then, that these two decisions conceive the confes-sion obtained as something different than a guilty plea. The confession agree-ment cannot just be a procedural excuse to dispose of a criminal case. It hasto provide elements of proof and be credible enough to support a criminalconviction. Otherwise, the court has to produce additional elements of proofat trial.

Nevertheless, there has been a second tendency within the German crimi-nal procedure to interpret these bargained confessions as the equivalent of aguilty plea.205 An example of this tendency is the decision of June 10, 1998,issued by a different senate of the Federal Supreme Court.206 In this case,after the reading of the information (indictment) at the beginning of thetrial, the trial court asked the defendant whether the information was cor-rect, to which he answered afªrmatively. Based on this simple acceptance ofthe information, the trial court convicted the defendant, and sentenced himto eight years in prison. The BGHSt, Second Chamber, upheld the deci-sion.207

199. See id. at 419 (discussing the principles of culpability or mens rea (das Schuldprinzip) and equality

or equal protection (der Gleichheitssatz) as also constraining the Absprachen).200. Id. The principle of investigation (der Ermittlungsgrundsatz), also called the principle of the mate-

rial truth or the inquisitorial principle, means that the judge makes her own investigation of the case.This principle is established in the German Criminal Procedure Code (StPO) in sections 155 Nr. 2 and244 Nr. 2. See Roxin, supra note 66, § 15 at 95.

201. BverfGE (Judgment of Jan. 27, 1987), quoted in Volker Gallandi, 1987 Absprachen im Strafprozess419, 419.

202. BGHSt 43, 195 (F.R.G.).203. Id. at 204.204. Id. For an analysis of this decision and the remaining requisites imposed on the agreements, see

Roxin, supra note 66, § 15 at 96–97.205. This shows how the translation of bargains to the German criminal procedure may eventually

transform the meaning of particular legal concepts such as the concept of confession, if not the model ofthe ofªcial investigation as a whole.

206. Bundesgerichtshof [BGH] [Federal Court of Justice] (Judgment of June 10, 1998), quoted in An-gabe des Grundes für den Ausschluss der Öffentlichkeit, 1999 NStZ 92, 92.

207. Id. at 93.


The interpretation of what constitutes this bargained confession is at thecenter of a discussion between two different conceptions of the criminal pro-cess. According to the ªrst conception, which has been predominant inGermany for a long time, the criminal procedure is an investigation con-ducted by the judge and the prosecutor to determine the truth. In this con-text, the bargained confession can be adapted to this procedural structure ofinterpretation and meaning, if it is interpreted and used as a tool to discoverthe truth.

According to the second conception, the criminal process is just an “as-sembly line” that processes criminal cases as quickly as possible.208 It is im-portant to remember, though, that the main manager of the “assembly line”in German plea bargaining is not the prosecutor, as it is in the U.S. model ofthe dispute, but rather the trial judge.209 Furthermore, although the trialjudge is the main manager, she still needs the cooperation of the defense andthe prosecution because the disposition of the case is only possible throughtheir participation. This means that if this tendency prevailed in Germany,its predominant structure of interpretation and meaning would be neitherthe model of the ofªcial investigation—nor the model of the dispute. Itwould be something different that I will provisionally call the “model of thejudge-manager based on cooperation.”210

In this judge-manager system, the role of the judge is to be neither a pas-sive umpire between the parties nor an active investigator. Rather, it is toassure that criminal cases are processed as quickly as possible. The judge,using his larger power, subtly (or not so subtly) presses the prosecution anddefense to collaborate in the swift disposition of the case.211 The prosecutor

208. The reference is, of course, to Herbert Packer’s famous Crime Control Model that conceives

criminal procedure as an “assembly line” and is opposed to the Due Process Model that conceives it as an“obstacle course.” See Packer, supra note 21. However, as I will explain, this German model differs fromPacker’s Crime Control Model, even if they share the goal of disposing criminal cases swiftly.

209. This active role of the judge distinguishes this tendency in German criminal procedure fromPacker’s Crime Control Model. Packer’s Crime Control Model presupposes the existence of the adversarialsystem, see id. at 157, and thus of a relatively passive judge. But since the judge is active in this Germanmodel, the German model differs from Packer’s.

210. The model of the judge-manager based on coordination presents similarities to tendencies inAmerican civil procedure of the judge being an active manager of the controversy rather than a passiveumpire as in the adversarial system. Among the seminal articles that describe some of these changes inU.S. civil procedure, see, for example, Abram Chayes, The Role of the Judge in Public Law Litigation, 89Harv. L. Rev. 1281 (1976); Owen M. Fiss, Foreword: The Forms of Justice, 93 Harv. L. Rev. 1 (1979); andJudith Resnik, Managerial Judges, 96 Harv. L. Rev. 374 (1982). For an analysis that challenges the viewthat these developments are entirely new to American civil procedure, see Theodore Eisenberg & StephenC. Yeazell, The Ordinary and the Extraordinary in Institutional Litigation, 93 Harv. L. Rev. 465 (1980). Foran analysis that questions whether these developments adequately describe contemporary complex U.S.civil litigation, see William B. Rubenstein, A Transactional Model of Adjudication, 89 Geo. L.J. 371

(2001).

211. Though most U.S. plea bargains happen through negotiations conducted exclusively between theprosecution and the defense, there are judges who, in certain cases, encourage the parties to reach anagreement in order to avoid the trial. In this sense, the system of the judge-manager based on cooperationmay also have some presence in the United States. Nevertheless, this practice is relatively minor amongU.S. plea bargains, at least within the regular criminal justice system, and the system of the dispute is


is conceived of as a collaborator with the judge who also contributes to therapid processing of criminal cases. The defense attorney is both an agent ofthe defendant who has to get the best possible outcome for his principal aswell as a professional legal actor who has a duty to collaborate with the otherprofessional legal actors in the swift disposition of cases.212

This does not mean, however, that the truth does not play any role in thismodel. As in the model of the dispute and the model of the ofªcial investi-gation, nobody wants to convict an innocent person or to acquit a guiltyone. However, ªnding the truth is a goal at least as important as the expedi-ent disposition of the case, and the procedure is thus not structured as aninvestigation.213

The actual importance of the system of judge-manager based on coopera-tion in current German criminal procedure poses an empirical question be-yond the scope of this Article.214 The points that need to be emphasized hereare that there are already indications that this system has a certain presence,at least in an incipient form, in contemporary German criminal procedureand that this system is different from both the model of the dispute and themodel of the ofªcial investigation. Also, to a certain extent, it seems to havebeen developed as a consequence of the translation of the bargained agree-ments in that country during the 1970s, and as a consequence of the nego-tiations that the legal translators had to make with the pre-existing inquisi-torial structure of interpretation and meaning, predominant internal dispo-sitions among legal actors, and the distribution of procedural powers.

To summarize, during the 1970s, the inquisitorial system, with its modelof the ofªcial investigation, was the prevailing procedural structure of inter-pretation and meaning in Germany, and most actors had incorporated thisstructure into their internal dispositions. Due to external societal changesthat increased the number and complexity of criminal cases, German legal

still predominant in this country. These kinds of practices are more extensive in U.S. civil cases. See, e.g.,Resnik, supra note 210.

212. On the redeªnition of the role of the defense attorney in the German agreements, see Weigend,supra note 191, at 779.

213. The conception of truth in the model of the judge-manager based on cooperation is more relativeand consensual than in the inquisitorial system. If the prosecutor, the defense, and the court agree thatevents transpired in a certain way—i.e., through agreed confessions—how events actually happenedbecomes less relevant. In this sense, the conception of truth in this model presents similarities to theconception of truth in the adversarial system. Therefore, even if the translation of confession agreementsis not pushing German practices in the direction of the model of the dispute, the Absprachen have thepotential to introduce a conception of truth closer to that of the adversarial system. If this potential werefulªlled, the result would be an Americanization of the German system regarding the conception of trutheven if there were no Americanization regarding the model of the dispute. This shows that Americaniza-tion is not an all-or-nothing game. The introduction of the German Absprachen could eventually createnew divergences with the American model of the dispute because of the advancement of the model of thejudge-manager based on cooperation and, at the same time, a convergence with the American adversarialsystem on a more relative and consensual conception of truth.

214. I leave for future work a more detailed analysis of the model of the judge-manager based on co-operation, as well as the question of whether it may be developing in other jurisdictions besides Ger-many.


actors received strong incentives to change their attitudes toward bargainsand agreements with the defendant. Nevertheless, plea bargaining suffered adeep transformation when it was translated from the adversarial structure ofmeaning to an inquisitorial setting because of the pre-existing inquisitorialstructure of interpretation and meaning, predominant individual disposi-tions and structure of procedural power, which neutralized, to a certain ex-tent, its potential Americanization or adversarialization effect on the Ger-man criminal process.

However, this new mechanism did present a challenge to the prevailinginquisitorial structure of meaning based on the model of the ofªcial investi-gation. The seriousness of this challenge is a question that has yet to be de-termined through empirical studies. However, since the introduction of theAbsprachen, there has been a struggle between the model of the ofªcial inves-tigation to adapt the new mechanism to its own logic and the model of thejudge-manager based on cooperation to interpret this mechanism in a differ-ent direction. This struggle should be understood not only in abstract termsas a clash between two structures of interpretation of meaning. It should alsobe understood as a struggle between legal actors with internal dispositionsthat correspond to these structures of meaning.215 The interactions betweenthese two structures of meaning—and other structures that are present inGerman criminal procedure and society—may also have an impact on otheraspects of German criminal justice practices, such as the organization ofhuman and material resources, legal ethics, etc. An analysis of these changesshould be included in a larger study on the transformations produced by thetranslation of plea bargaining into Germany.

From the perspective of the Americanization thesis debate, the Germanexample shows that, due to the transformation that it suffered in its transla-tion and the interactions that followed, the German Absprachen have notAmericanized German criminal procedure, in the sense that it has not intro-duced a cultural conception of criminal procedure as a dispute betweenprosecution and defense before a passive decision-maker. Rather, due to thislegal translation, this procedure may be at a crossroads between its tradi-tional inquisitorial model of criminal procedure and a model that is neitherinquisitorial nor adversarial.

VIII. Italian “Plea Bargaining”

Compared to the German Absprachen, Italian plea bargaining, or patteg-giamento, has been much more faithful to the American model, even if it pre-sents substantial differences from the American practice. Italian legal trans-lators had much more power than German translators in advancing theirreform, and were able to introduce Italian plea bargaining as part of deeper

215. The struggle may occur not only between individual legal actors but also within individual legal

actors who possess internal ambivalence about both systems.


and broader criminal procedure reforms inspired by the American adversar-ial system. Thus, Italian reformers had to compromise much less than Ger-man translators, and they had much more power to change the pre-existinginquisitorial structure of meaning, distribution of procedural powers, etc.Since these adversarial reforms have had substantial political support, theyhave moved the Italian system in the direction of the American adversarialsystem much more than any other civil law jurisdiction. However, even inthe Italian system today, many legal actors are still moved by a predomi-nantly inquisitorial set of internal dispositions. Furthermore, the inquisito-rial structures of interpretation and meaning and procedural power remainpresent.

In 1989, Italy adopted a new criminal procedure code that replaced theRocco Criminal Procedure Code enacted during Mussolini’s regime.216 Thisnew code represented the most serious attempt to transfer adversarial crimi-nal procedures into an inquisitorial jurisdiction since 1791, when the Frenchattempted to import the English system during the heat of the Revolu-tion.217 The Italian reform was justiªed both on due process and efªciencygrounds, but had a predominantly due process inspiration.218 Americancriminal procedure was the main model for these adversarial reforms becauseof the prestige that the U.S. legal system in general, and U.S. criminal pro-cedure in particular, enjoyed among a substantial number of legal actorssince the end of the Second World War.219

The adversarial reforms can be described and analyzed by using the twopairs of models developed in Part III: the coordinate model vs. the hierarchi-cal model and the model of the dispute vs. the model of the ofªcial investi-gation.220

Regarding the ªrst pair, the reformers tried to import the oral case-management technique of the coordinate model by reducing the importanceof the evidence collected during the pre-trial phase in the distinctively in-quisitorial written dossier.221 They did so by ordering the preparation of aspecial written dossier for the trial so that the trial court did not haveknowledge of the testimonies collected during the pre-trial phase prior tothe trial,222 as well as by introducing rules of evidence such as hearsay that

216. In that sense, the introduction of the code could be analyzed as part of the transition to democ-

racy that Italy began after the Second World War.217. The classic account of this importation is Esmein, supra note 63, at 399–480.218. See Vassalli, supra note 115.219. See Grande, supra note 32, at 230–32.220. For analyses in English of these reforms, see Amodio & Selvaggi, supra note 115; Stephen P. Frec-

cero, An Introduction to the New Italian Criminal Procedure, 21 Am. J. Crim. L. 345 (1994); William T.Pizzi & Luca Maraªoti, The New Italian Code of Criminal Procedure: The Difªculties of Building an Adversar-ial Trial System on a Civil Law Foundation, 17 Yale J. Int’l L. 1 (1992).

221. See, e.g., Amodio & Selvaggi, supra note 115, at 1217.222. C.p.p. art. 431 (1989) (Italy).


prevented pre-trial testimony from being admitted at trial, save for im-peachment purposes.223

Regarding the second subsystem, Italian legal reformers have introducednumerous reforms to advance the model of the dispute within the Italiansystem. First, the parties are now supposed to develop the fact-ªnding proc-ess, both at the pre-trial phase and at trial. This has eliminated the prelimi-nary investigation judge (giudice istruttore) as the fact-gatherer during thepre-trial phase,224 replacing him with the prosecutor and defense attorney,who can now conduct their own investigations.225 It has also entailed theorganization of the trial into a case for the prosecution and a case for the de-fense,226 and introduced direct and cross-examination as the way to interro-gate witnesses at trial.227 Besides these and other reforms, the Italian Codeincludes a rule establishing that the judge can only order the production ofevidence requested by the parties; only in exceptional circumstances can thejudge do so sua sponte.228

Second, the reform has also introduced a number of simplifying and con-sensual mechanisms through which the prosecution and defense can bargainand agree to avoid the regular proceedings.229 The introduction of thesesimplifying mechanisms has been justiªed on efªciency grounds.230 How-ever, they are also part of this broader reform of Italian criminal procedure inwhich the cultural inºuences of the American model—as a system that hassymbolized democratic values and efªciency in dealing with crime—hasplayed an important role. These mechanisms have included the applicazionedella pena sulla rechiesta delle parti—application of the punishment upon therequest of the parties231—that is usually called Italian plea bargaining, orsimply patteggiamento, which means “bargain” in Italian.232

223. C.p.p. arts. 500, 503 (1989) (Italy). The inquisitorial reaction against these ideas—combined

with a delicate political situation in which two of the main anti-Maªa judges had been assassinated—came a few years after the reform when the Italian Constitutional Court, in its decisions 254/1992 and255/1992, held unconstitutional art. 513.2 and arts. 500.3 and 500.4. Decision 255/1992, supra note163; Corte cost., 18 mag. 1992, n.254, 103 Racc. uff. corte cost. 1992, 809. Nevertheless, in 1999, theItalian Parliament introduced a new Article 111 to the Italian Constitution, backing the original direc-tion of the 1989 reforms toward an increasing role for the oral case-management techniques of the coor-dinate system.

224. See Vassalli, supra note 115, at 6.225. Regarding the power of the prosecution to make the pre-trial investigation, see C.p.p. art. 358

(Italy). The investigatory powers of the defense were included in Law 271/1989, supra note 164, and laterincorporated in the Criminal Procedure Code, articles 391-bis-decies by Law 397/2000, supra note 164.

226. C.p.p. art. 496.1 (Italy).227. C.p.p. art. 498 (Italy).228. C.p.p. arts. 190, 493.1, 495.1, 507 (Italy).229. C.p.p. arts. 438–64 (Italy).230. See, e.g., Vassalli, supra note 115, at 7.231. A forerunner of this institution was the applicazione di sanzioni sostitutive su richiesta dell’imputato

introduced in 1981 by L.vo 24 nov. 1981, n.689, Racc. Uff. 1981, vol. 13, 3657, published in Gazz.Uff., 30 nov. 1981, n.329. The scope of this mechanism was very limited because it only applied tooffenses punishable by up to three months in prison. See, e.g., Pizzi & Maraªoti, supra note 220, at 22.

232. Besides the patteggiamento, the 1989 reform introduced two other mechanisms that may be com-pared to plea bargaining because they included potential negotiations between prosecution and defense,


One crucial difference between the introduction of the German Absprachenand the patteggiamento is that the latter was introduced by statute and as partof a reform that had substantial political support. In this sense, the transla-tors233 had much more freedom to decide the extent to which they wanted tobe faithful to U.S. plea bargaining. Moreover, the Italian reformers had avery sophisticated knowledge of the U.S. criminal justice system,234 whichcoupled with their legislative freedom, allowed them to design a mechanismthat was much more similar to U.S. plea bargaining than the German Ab-sprachen while still taking into account the pre-existing inquisitorial struc-ture of meaning, internal dispositions, and structure of power.

Under the patteggiamento, the defense and the prosecution can reach anagreement about a sentence and request that it be imposed by the judge.235

Through this agreement, the regular sentence can be reduced by up to onethird if the reduced sentence will not exceed ªve years of imprisonment.236

and required the consent of the defendant to be applied. The ªrst was the procedimento per decreto. C.p.p.arts. 459–64 (Italy) (similar to the German penal order analyzed in Herrmann, Bargaining Justice, supranote 183). The second is the giudizio abbreviato. C.p.p. arts. 438–43 (Italy) (in which the defendant waivesthe right to trial and accepts being tried in the preliminary hearing). In exchange, he receives a reductionof one third of the regular sentence. This mechanism initially required the consent of the prosecution forits application, and thus it opened the door for negotiations between prosecution and defense. Neverthe-less, since then, ªrst the courts and ªnally the Parliament have limited the consent of the prosecution as arequirement for its application. For an analysis of this process, see Elena Maria Catalano, Il GiudizioAbbreviato, in Giudice Unico e Garanzie Defensive 117 (Ennio Amodio & Novella Galantini eds.,2000). For space limitations, I will not analyze these two mechanisms in the main text but they shouldbe included in a larger work.

233. The Italian Criminal Procedure Code was drafted by a Commission designated by d.m. (ministe-rial decree) of March 3, 1987, and was composed by nine law professors: Giandomenico Pisapia (Presi-dent), Delªno Siracusano (Vice-president), Ennio Amodio, Vincenzo Cavallari, Mario Chiavario, OresteDominioni, Vittorio Grevi, Guido Neppi Modona, and Mario Pisani; eight judges: Giancarlo Caselli,Enrico Di Nicola, Liliana Ferraro, Giuseppe La Greca, Giorio Lattanzi, Enrnesto Lupo, Vittorio Mele, andPiero Luigi Vigna; and a lawyer: Giusepe Frigo. See Vassalli, supra note 115, at 3.

234. For example, Prof. Ennio Amodio was member of the Ministerial Commission that designed thereforms and is, at the same time, a sophisticated comparativist. See, e.g., Ennio Amodio, Il Modello Accu-satorio Statunitense e il Nuovo Processo Penale Italiano: Miti e Realtà della Giustizia Americana, in Il Processo

Penale negli Stati Uniti d’America VII (Ennio Amodio & M. Cherif Bassiouni eds., 1988).235. The patteggiamento is regulated in articles 444–48 of the Italian Criminal Procedure Code. For a

detailed analysis of these articles, see Franco Cordero, Procedura penale 960–73 (5th ed. 2000)(Professor Cordero, however, does not include in his analysis the reforms introduced by law 134/2003,which I analyze infra, note 236.). In the original code, the parties could request the application of themechanism until the opening of the trial. The law n. 479/1999 limited this term to the end of the pre-liminary hearing (with exceptions I will not analyze here). L.vo 16 dic. 1999, n.479, Racc. Uff. 1999,vol. 13, 6496, published in Gazz. Uff., 18 dic. 1999, n.296. The more recent reform to the patteggiamentointroduced by law 134/2003 has not changed this aspect, and as a general rule, the parties can request theapplication of this mechanism only until the end of the preliminary hearing. See C.p.p. art. 446.1 (Italy).The parties cannot request the application of the patteggiamento after they have made their conclusions inthe preliminary hearing.

236. C.p.p. art. 444.1 (Italy). The limitation of up to ªve years of imprisonment was introduced bystatute in June 2003. See L.vo 12 giu. 2003, n.134, art. 1, 1962, published in Gazz. Uff., 14 giu. 2003,n.136 [hereinafter Law 134/2003] (if the agreed sentence is between two and ªve years of imprisonment,certain offenses, such as terrorism and organized crime, are excluded). Before this last reform, the bar-gained sentence could not exceed two years of imprisonment—after the sentence reduction—and thuswas limited to minor crimes. The two justiªcations given to extend the application of the patteggiamentoto more offenses were reducing the caseload of the criminal justice system and limiting the use of full


If, upon examining the case’s dossier, the judge does not ªnd any sufªcientreasons to acquit the defendant237 and considers the charge and sentence tobe proportional to the offense, she will apply the requested punishment.238

Though much more similar to U.S. plea bargaining than the German Ab-sprachen, there are several differences between the patteggiamento and Americanplea bargaining that reºect, in part, the ambivalence of Italian translatorstoward the very mechanism they introduced. First, the patteggiamento is morelimited in scope and less ºexible. It can only be applied in cases where thesentence does not exceed ªve years of imprisonment after sentence reduction,and the sentence reduction bargained for by the parties cannot be greaterthan one-third of the regular sentence for the case.239 Furthermore, at least asit was originally conceived by the reformers, the bargain can only concernthe sentence, not the charge or charges.240 The decision of the translators toestablish these limitations likely reºects two different concerns. On the onehand, the translators were ambivalent toward this bargaining mechanismbecause they were aware of the due process problems241 that plea bargainingwas generating in the United States.242 On the other hand, they were con-scious that legal actors with an inquisitorial set of internal dispositions (andaccustomed to an inquisitorial distribution of procedural powers) might re-sist the translated institution. Thus, they decided to introduce it only forminor offenses and to limit the practice to sentencing bargains.243 Also, evenif a recent reform has expanded the number of offenses that can be negoti-

adversary trials to the most serious offenses. See Progetto di Legge—N. 1488, available at http://www.camera.it/_dati/leg14/lavori/stampati/sk1500/relazion/1488.htm (last visited Nov. 25, 2003) (on ªle with the HarvardInternational Law Journal) (initially proposing to extend the pattegiamento to sentences of up to four yearsof imprisonment after the one-third reduction). There have been suggestions that the statute was intro-duced for political reasons because its art. 5 (unrelated to the extension of the patteggiamento to agree-ments on sentences of up to ªve years of imprisonment) established that in pending trials, the defendantcan require a 45-day continuance for thinking about requesting the application of the patteggiamento. Thiscontinuance was used by Cesare Previti—a person close to Silvio Berlusconi—in the Sme case, in whichSilvio Berlusconi himself is a defendant, even if the case cannot proceed while he is in ofªce. See, e.g.,Berlusconi, processo sospeso sul “lodo” atti alla Consulta, La Repubblica (Italy), June 30, 2003, at Politica,available at http://www.repubblica.it/online/politica/smestralciocinque/sospeso/sospeso.html (last visitedNov. 26, 2003) (on ªle with the Harvard International Law Journal). Thus, it has been insinuated thatthis art. 5 was actually introduced to block this case. See, e.g., Vittorio Grevi, Doppio Blocco, DS Milano,available at http://www.dsmilano.it/Pressroom/2003/07/cor3_0701_grevi-doppio-blocco.htm. (last vis-ited Nov. 25, 2003) (on ªle with the Harvard International Law Journal).

237. C.p.p. arts. 444.2, 129 (Italy).238. C.p.p. art. 444.2 (Italy). If the agreed sentence is for up to two years of imprisonment, the judge

can replace it with a substitute punishment such as semi-detention (for sentences of up to two years ofimprisonment), freedom under surveillance (for sentences of up to one year of imprisonment), or a ªne(for sentences of up to six months of imprisonment). See Law 134/2003, supra note 236, art. 4.1(a).

239. C.p.p. art. 444.1 (Italy).240. See, e.g., Pizzi & Maraªoti, supra note 220, at 22. However, there are indications that charge bar-

gains may have been introduced by practitioners and accepted by the courts. See Cass. pen., sez. cin.,7 ott. 1998, n.12743.

241. Recall that the reform had a clear garantista—due process—inspiration.242. On these criticisms in the United States, see the bibliography cited supra note 170.243. Charge bargains usually compromise the material truth more than sentencing bargains. Thus,

the latter would be more acceptable from an inquisitorial perspective.


ated through the patteggiamento, the practice still does not apply to all of-fenses.244

Second, in the patteggiamento there is no guilty plea or explicit admissionof guilt by the defendant.245 By requesting the application of the sentence,the defendant waives his right to a trial and may be implicitly admitting hisguilt. However, the judge can still decide to acquit the defendant after ex-amining the evidence collected in the written dossier and before acceptingthe agreement.246 This decision not to introduce an explicit admission ofguilt with the patteggiamento is another reºection of due process concerns.The drafters feared that an admission of guilt would undermine the pre-sumption of innocence guaranteed to all defendants in the Italian Constitu-tion.247 In this sense, the absence of an explicit admission of guilt makes theItalian pattegiamento more similar to U.S. negotiations about nolo contenderethan about guilty pleas.

Third, when the prosecutor does not accept an agreement with the defen-dant, the latter can ask the judge at the end of the trial to examine the rea-sons given by the prosecutor to reject such an agreement, and to give himthe beneªt of the one-third reduction of the sentence.248 This differencereºects the inºuence of the model of the ofªcial investigation on the internaldispositions of the translators.249 As explained above, granting beneªts tothe defendant in exchange for admissions of guilt—in this case, implicitadmissions of guilt—is not unknown in the Italian structure. What is for-eign, though, is the negotiation of a sentence that dispossesses the judge ofher powers over it. Even if the patteggiamento generally empowers the partiesagainst the powers of the judge, this rule afªrms that the judge still con-serves a certain quantum of power in the application of this mechanism, andthat the mechanism is not only a bargaining procedure but also a beneªtthat the judge may concede to the defendant.

Finally, the sentence pronounced under this procedure250 does not haveany effect on civil and administrative proceedings,251 which again makes the

244. On this reform, see C.p.p. 444.1 (Italy).245. See, e.g., Pizzi & Maraªoti, supra note 220, at 23.246. C.p.p. arts. 444.2, 129 (Italy).247. See Pizzi & Maraªoti, supra note 220, at 23.248. C.p.p. art. 448.1 (Italy).249. As Pizzi & Maraªoti have noted,

The intent of the Italian Code is to make sentence reduction available to all defendants who wish toplea bargain, whether or not the prosecutor agrees. This arrangement reºects the traditional civillaw distrust of prosecutorial discretion and commitment to uniform treatment of defendants—thatdefendants would receive different sentences simply because of a prosecutor’s whim is anathema tocivil law.

Pizzi & Maraªoti, supra note 220, at 22–23250. The verdict based on a patteggiamento is considered “equivalent to a verdict of guilt” (C.p.p. art.

445.1-bis (Italy)), but it is not like the verdict issued after the regular trial. This explains the more lim-ited effects that the verdict and sentence based on a patteggiamento have and shows, again, the distrust andambivalence that the reformers had towards this mechanism.

251. C.p.p. art. 445.1-bis (Italy). The only exception to this rule is that the bargained sentence consti-tutes res judicata for disciplinary proceedings before the public authorities. See C.P.P. arts. 445.1-bis,


patteggiamento closer to the U.S. nolo contendere than to a plea bargain. Thislimitation of the legal effects of the mechanism may also reºect the ambiva-lence the translators had towards it, both because of due process concernsand the inºuence of the model of the investigation on their internal disposi-tions.

Nevertheless, despite their ambivalence, the Italian translation of plea bar-gaining is faithful to the original American mechanism in that it is a proce-dure through which the prosecution and the defense can actively negotiatesentences and the judge is assigned a relatively passive position. In thissense, the adoption of the patteggiamento has had, from the very beginning, apotential Americanization effect; if accepted and internalized by Italian legalactors as a negotiating mechanism, and accompanied by the other adversarialreforms, the patteggiamento has the potential to move Italian criminal proce-dure practices toward the model of the dispute, replacing the predominantinquisitorial internal dispositions of the legal actors and the structure of in-terpretation and meaning, and shifting the distribution of procedural power.

It is not surprising, then, that there has been some resistance to this newprocedure, both among legal commentators and the courts.252 The most im-portant reaction of the courts thus far has been the Italian ConstitutionalCourt’s 313/1990 decision, issued the year after the new Italian Code intro-duced the patteggiamento.253 In this decision, after repeatedly stating that thejudge’s power of control over the agreement was not just a formality,254 theConstitutional Court held that art. 444.2 of the Italian Criminal ProcedureCode, which regulated the patteggiamento, was unconstitutional because itdid not expressly give the judge the power to control the congruence be-tween the sentence agreed upon by the parties and the seriousness of theoffense, and thus deprived the judge of the power to enforce art. 27.3 of theItalian Constitution, which establishes that the goal of punishment is therehabilitation of the convicted person.255 Therefore, this decision attemptedto reafªrm and increase the powers of the judge against the parties regardingthe pattegiamento.

Despite this reaction, Italian legal actors have used the patteggiamentoquite extensively. Between 1990 and 1998, the number of cases disposedthrough this procedure before the pretura (misdemeanor jurisdiction) wasbetween 17 and 21%; and, in the case of the tribunale (the jurisdiction for all

653.1-bis (Italy).

252. Among the criticisms from commentators, see Ferrajoli, supra note 66; and Ferrua, supra note179. Both cited criticisms have their origin in their defense of the substructure of the ofªcial investiga-tion because both point out how bargaining mechanisms compromise the real truth. Their critique,however, is inspired by due process concerns because both commentators understand the truth—and thecognitive role of the judge—as a protection for the accused. For a very sophisticated defense of this (ga-rantista) conception of the criminal process, see Ferrajoli, supra note 66.

253. Corte cost., 26 giu. 1990, n.313, 96 Racc. uff. corte cost. 1990, 89.254. See, e.g., id. at 97–98.255. Id. at 102–05.


the crimes except the most serious ones) was between 34 and 42%.256 Nowthat the application of the pattegiamento has been extended to more seriousoffenses,257 the Italian actors are likely to use it even more widely.

This indicates that the model of the dispute, understood as a structure ofinterpretation and meaning, has been accepted and internalized, at least to acertain extent, by a substantial number of Italian legal actors. Furthermore, as aconsequence of the reform to the Italian criminal procedure which began in1989, there are other indications that the model of the dispute has a substantialpresence in Italian criminal proceedings. These include the possibility of theprosecution and the defense doing their own pre-trial investigation, the organi-zation of the trial into the case of the prosecution and the case of the defense,and the introduction of direct and cross-examination, as mentioned above.

However, this does not mean that the model of the ofªcial investigationlacks a substantial presence in Italian criminal procedure and in the individ-ual dispositions of a signiªcant number of legal actors. Many Italian judges,for instance, still believe they have a duty to ensure that the truth prevailsand thus participate actively in the fact-ªnding process.258 Furthermore,both Italian prosecutors and judges remain members of the judiciary and aretrained together. Thus, many Italian prosecutors still consider themselves,and are considered by judges, to be magistrates whose role is to investigatethe truth, and not as mere parties to a dispute.259

Analyzing in detail these particular struggles between the dispute modeland the ofªcial investigation model is beyond the scope of this Article. Thepurpose of this Part has been to show how U.S. plea bargaining has beentranslated into the Italian system and the kinds of transformations this transla-tion may be producing. From the perspective of the debate about the Ameri-canization thesis, the patteggiamento and other reforms have indeed advancedthe model of the dispute as a structure of interpretation and meaning. In noother civil law country has this structure arguably attained such an impor-tant status. In this sense, the development of the 1989 reform seems to haveput Italy at a cross-roads between the adversarial and inquisitorial systems.

IX. Argentine “Plea Bargaining”

Of the four systems examined in this Article, Argentina’s plea bargainingsystem has been the most faithful translation of U.S. plea bargaining. How-ever, this reform represents one of a very small number of Argentine reformsin the direction of the adversarial system. Given that the pre-existing in-

256. For statistical information for 1990 and 1991, see Istituto Centrale de Statistica, Statis-

tiche Giudiziarie; for 1992 to 1998, see Istituto nazionale di statistica, Statistiche Gi-

udiziarie Penali. I do not include statistics from 1989 because the total number of sample cases ismuch smaller for that year than for the other years.

257. See supra note 236 on the Legge, 12 giugno 2003, n.134, art. 1.258. See, e.g., Grande, supra note 32, at 250–51.259. About these and other resistances resulting from the structure of the ofªcial investigation, see

Grande, supra note 32.


quisitorial structure of interpretation and meaning, internal dispositions,and structure of procedural power of the Argentine system remain predomi-nant, it is highly unlikely that this adversarial reform alone will American-ize Argentine criminal procedure in the sense of moving its criminal proce-dure practices and culture in the direction of the model of the dispute.

In June of 1997, Argentina incorporated into its Federal Code of CriminalProcedure260 the so-called procedimiento abreviado261 in order to speed up pro-ceedings and reduce the caseload of trial courts.262 This mechanism seems tohave been adopted mainly for pragmatic reasons rather than as a result ofdeep cultural inºuences of the American system on the Argentine system,though these inºuences indeed have been present.263 In any case, Americanplea bargaining has been an important source for the development of theArgentine practice. According to the procedimiento abreviado, the prosecutionand the defense can reach an agreement about the sentence at any time be-tween the production of the information (indictment) at the end of the pre-trial phase and the determination of the date for trial.264 This negotiatedsentence cannot be greater than six years of imprisonment.265 As part of theagreement, the defendant must admit to the offense and his participation init as described in the indictment.266 The trial court can reject the agreementif it considers the production of additional evidence necessary, or if it fun-damentally disagrees with the charges.267 However, if the trial court acceptsthe agreement, it must reach a verdict based on the evidence collected in thewritten dossier.268 The trial court can still acquit the defendant, but if con-

260. Argentina is a federal state consisting of 23 provinces. There is one Code of Criminal Law for all

these jurisdictions. But each province has its own code of criminal procedure and its own system ofcourts. The Federal Code of Criminal Procedure regulates the proceedings of the federal courts and of thecourts of the city of Buenos Aires. In this Article, I will concentrate on these proceedings.

261. The procedimiento abreviado is also called juicio abreviado. See Cód. Proc. Pen. art. 431 (Arg.).262. Besides these two goals, other goals that were considered during the hearings of the Argentine

Congress on this institution were reducing the number of pre-trial detentions, reducing the cost ofcriminal trials, rationalizing the use of criminal procedure resources, and giving defendants the opportu-nity to receive a reduction in their potential sentences. See “Wasylyszyn, M.A.,” Trib. Oral Crim. no. 20, 8aCuadernos de Doctrina y Jurisprudencia Penal 623, 628 (1998) (describing the intervention bythe representative José I. Cafferata Nores in the parliamentary debate of October 23, 1996). See also José

I. Cafferata Nores, Cuestiones actuales sobre el proceso penal 143 (2d ed. 1998).263. The American criminal justice system has had a deep cultural inºuence on the precedents of the

Argentine Supreme Court concerning constitutional rights in criminal procedure (searches and seizures,the right against compelled self-incrimination, exclusionary rules, fruit of the poisonous tree doctrine,etc.). For an analysis of these types of decisions by the Argentine Supreme Court, see generally Alejan-

dro Carrio, Garantías constitucionales en el proceso penal (3d ed. 1997). Concerning theinºuence of U.S. Supreme Court precedents on the Supreme Court of Argentina, see Carlos IgnacioSuárez Anzorena, Transnational Precedents: The Argentinean Case 26–40 (1998) (unpublished LL.M. thesis,Harvard Law School) (on ªle with the Harvard Law School Library). However, the American criminaljustice system has not been as inºuential as it was in the Italian case. The Italian system has importedvery substantial parts of American criminal procedure, rather than just selected pieces of it.

264. Cód. Proc. Pen. art. 431 bis. 1 (Arg.).265. Id.266. Cód. Proc. Pen. art. 431 bis. 2 (Arg.).267. Cód. Proc. Pen. art. 431 bis. 3 (Arg.).268. Cód. Proc. Pen. art. 431 bis. 5 (Arg.).


victed, the defendant’s sentence cannot exceed the length agreed to by theparties.269

Of all the legal translations analyzed in this Article, the Argentine pro-cedimiento abreviado seems to be the most faithful to U.S. plea bargaining, ormore speciªcally, to U.S. sentencing bargaining.270 The prosecution and de-fense have active roles in the negotiations about the sentence and the admis-sion of guilt while the tribunal’s role is basically limited to that of formalcontrol. In addition, the procedimiento abreviado includes an admission of guiltby the defendant similar to a guilty plea. Furthermore, because the partiescan agree on a sentence of up to six years, the procedimiento abreviado can beapplied to some, though not all, serious offenses.271

However, there are differences between the procedimiento abreviado and U.S.plea bargaining that reveal how the decisions made by Argentine translatorsand the pre-existing Argentine inquisitorial structure of interpretation andmeaning and distribution of procedural power transformed the practice.272

First, there is a temporal limitation as to when the parties can reach anagreement, making the entire procedure less ºexible.273 Second, in the pro-cedimiento abreviado, the judge can still acquit the defendant.274 The admis-sion of guilt, then, is not understood exactly as a guilty plea like that in theUnited States, but rather as a confession that may be disregarded by thecourt, exemplifying the inºuence of the pre-existing inquisitorial structureof meaning on the practice. Third, the trial court has to respect, as an upper

269. Id.270. U.S. plea bargaining was an important source for the development of the procedimiento abreviado.

See “Wasylyszyn, M.A.,” supra note 262, at 628–29.271. The procedure can be applied, for instance, to manslaughter, rape, and aggravated robbery.272. The two individuals who introduced the procedimiento abreviado in Argentina were Professors Julio

B. J. Maier (who included a similar mechanism in his drafted Federal Criminal Procedure Code of 1986,which the Congress did not enact, and in the Model Criminal Procedure Code for Ibero-America) andJosé I. Cafferata Nores (university professor, former minister of justice of the province of Cordoba andmember of the Federal House of Representatives when the procedimiento abreviado was enacted). Both ofthem conceived of the institution only for minor cases. For instance, in Article 371 of his drafted FederalCriminal Procedure Code, Prof. Maier conceived of it for handling cases where the sentence was notgreater than one year of imprisonment. Julio B.J. Maier, El Proyecto de Código Procesal Penal

de la Nación: Presentación de Julio B.J. Maier, exposición de motivos, texto completo del

proyecto 764 (Cuadernos de la Revista Doctrina Penal, Series No.1, 1987). They believed that thismechanism would be useful to handle minor cases faster, but they also distrusted it because they wereconscious that the mechanism implied a completely different conception of criminal procedure. Forinstance, Maier clearly saw that the mechanism implied a more consensual and relative conception oftruth. See Julio B. J. Maier, Mecanismos de Simpliªcación del Procedimiento Penal, in 8a Cuadernos de Doc-

trina y Jurisprudencia Penal, supra note 267, at 433, 435 (1998). The distrust of Maier and CafferataNores toward the institution was also based on due process concerns because they knew of the critiquesthat these kinds of agreements were generating in the United States and elsewhere. Despite their at-tempts to introduce the mechanism only for minor cases, the Argentine Congress has extended its appli-cation to relatively serious crimes.

273. See Cód. Proc. Pen. art. 431 bis. 1 (Arg.).274. Cód. Proc. Pen. art. 431 bis. 5 (Arg.) states that once the prosecution and defense have reached

their agreement, the court still has to decide its verdict. Thus, it is technically possible that the courtwill decide to acquit the defendant after the agreement. However, the law in action regarding this issue isthat courts usually issue a conviction.


limit, the sentence agreed upon by the parties.275 This limitation does notalways apply in U.S. sentencing bargaining, though most judges respect itde facto.276 Fourth, in a case with more than one defendant, the mechanismcan be applied only if accepted by each defendant.277 This shows distrusttoward the mechanism from a due process perspective.278

Apart from these differences, the procedimiento abreviado is a faithful trans-lation of U.S. plea bargaining. As such, it assumes the model of the disputein the sense that both parties control the case and that the judge has a rela-tively passive role. In this sense, the procedimiento abreviado could be consid-ered something of a Trojan horse for the model of the dispute in the Argen-tine inquisitorial model of the ofªcial investigation.279 If Argentine legalactors accept and internalize this mechanism, the reform could potentiallyAmericanize (or adversarialize) Argentine criminal procedure.

The model of the ofªcial investigation has always been predominant inArgentine criminal procedure.280 Thus, it is not surprising that a number ofcommentators strongly criticized the introduction of a mechanism so clearlyinspired by the model of the dispute.281 Nevertheless, legal practitioners,prosecutors, and judges have generally accepted and used it.282 For instance,in the ªrst semester of 2000, 22% of the cases before the misdemeanors trialcourts of the city of Buenos Aires were disposed of through the procedimientoabreviado; 52% of the cases before the crimes trial courts in the same juris-

275. Id.276. For a proposal to incorporate this limitation into the U.S. system, see Scott & Stuntz, supra note

170 at 1953–57.277. Cód. Proc. Pen. art. 431 bis. 8 (Arg.).278. The requirement that all defendants have to accept to participate in the agreements may have

been aimed at avoiding use of the procedimiento abreviado for investigative purposes, i.e., for making adefendant testify against another.

279. For an analysis of the potential effects of the incorporation of the procedimiento abreviado in Argen-tina, see Langer, supra note 23, at 124.

280. For a description in English of the Argentine federal criminal procedure, see Alejandro D. Carrio& Alejandro M. Garro, Argentina, in Criminal Procedure: A Worldwide Study 3 (Craig M. Bradleyed., 1999).

281. See, e.g., Francisco J. D’Álbora, El proceso penal y los juicios abreviados, in 8a Cuadernos de

Doctrina y Jurisprudencia Penal, supra note 262, at 457, 463 (1998); Leopoldo H. Schiffrin, Corsi ericorsi de las garantías penales en la Argentina, in 8a Cuadernos de Doctrina y Jurisprudencia Penal,

supra note 267, at 481, 484 (1998). As in the Italian example, part of this critique was also based on dueprocess concerns because a number of commentators have recognized the due process criticism that pleabargaining has received in the United States.

282. The resistance against the procedimiento abreviado has mainly come from individual members of afew trial courts. In Argentina, trial courts for medium and serious offenses are composed of three judges.There have been some dissenting opinions from members of these courts stating that the procedimientoabreviado is unconstitutional because a real trial is a necessary condition for a criminal conviction, whichcannot be replaced by an agreement between prosecution and defense. See, e.g., “Dos Santos Amaral, M.,”Trib. Pen. Económico no. 3, 8a Cuadernos de Doctrina y Jurisprudencia Penal, supra note 262, at613 (1998); “Wasylyszyn, M.A.,” supra note 262; “Osorio Sosa, A.,” Trib. Oral Crim. no. 23, 8aCuadernos de Doctrina y Jurisprudencia Penal, supra note 262, at 636 (1998). In addition, sometrial courts have also held the procedimiento abreviado unconstitutional. See, e.g., “Yunez, R.D.,” Trib. OralCrim. Fed. [1999-C] L.L. 335.


diction were disposed of in the same manner.283 One of the main reasons forthis widespread acceptance likely comes from strong external incentives:between 1990 and 2000, the number of offenses that entered into the crimi-nal justice system of the city of Buenos Aires increased from 61,203 to191,755 (313%).284 In a situation like this, almost any mechanism to dis-pose of criminal cases quickly would be welcome, even if it existed in ten-sion with the prevailing structure of interpretation and meaning, internaldispositions, and distribution of procedural powers.

Besides the introduction of the procedimiento abreviado, there have been twoother tendencies that could potentially move Argentine criminal procedurein the direction of the adversarial model of dispute. First, the rule of com-pulsory prosecution did not apply at trial for several years.285 Second, since1994, there has been a very clear institutional distinction between prosecu-tors and judges that may contribute to differentiating both roles within thecriminal process, a condition that makes the model of the dispute possi-ble.286

Nevertheless, it is important not to overemphasize the signiªcance of thesechanges, including the introduction of the procedimiento abreviado. These havebeen practically the only reforms introduced that could potentially lead theArgentine legal system in the direction of the model of the dispute. In therest of Argentine criminal procedure practices found in the federal jurisdic-tion and in the jurisdiction of the city of Buenos Aires, the inquisitorialstructure of interpretation and meaning, internal dispositions, and distribu-tion of procedural power remain predominant.287 For instance, the model ofthe ofªcial investigation is still quite evident in the role played by the

283. Ministerio Público Fiscal de la República Argentina, Procuradoría General de la

Nación, Informe Anual al Congreso de la Nación 2000 297–98, 285–86 (2000). These statisticsrefer to the trial phase. In the pre-trial phase, this practice is totally insigniªcant, hence, coverage of thepre-trial phase is not included here.

284. See Evolución de Hechos Delictuosos—1990 a 1999—Total País y Por Provincia, Ministerio deJusticia, Seguridad y Derechos Humanos de la Nación Argentina at http://www.jus.gov.ar/polcrim/InfoDecada/infosnic2C33A.pdf (last visited Dec. 2, 2003) (on ªle with Harvard International LawJournal).

285. See, e.g., “Tarifeño, F.,” CSJN [1995-B] L.L. 32; “García, J.A.,” CSJN [1995-B] L.L. 31; “Catto-nar, J.P.,” CSJN, [1996-A] L.L. 67. For a brief analysis of these decisions, see Julio B.J. Maier & MáximoLanger, Acusación y Sentencia, in 1996/B Nueva Doctrina Penal 617 (1996). The Argentine SupremeCourt, however, reversed its decisions on these matters, and the rule of compulsory prosecution nowapplies at trial. See “Marcilese, P.J.,” CSJN [2002-E] L.L. 719.

286. See Const. Arg. art. 120, amended by Law No. 24.309, Aug. 23, 1994, B.O. (Arg.) (establish-ing that the Ofªce of the Attorney General is part of neither the Executive power nor the Judiciary, but aseparate branch of government).

287. Several other Argentine jurisdictions have introduced additional adversarial reforms in theircriminal procedures during the 1990s, including the Province of Buenos Aires (1997) and the Province ofCórdoba (1991). However, these reforms are beyond the scope of this Article. These reforms have elimi-nated the preliminary investigation judge in most cases and put the prosecutor in charge of the investiga-tion during the pre-trial phase, see, e.g., Fabricio Guariglia & Eduardo Bertoni, supra note 117, at 66–67.However, even in these jurisdictions, the model of the ofªcial investigation is still very predominant,and, thus, these reforms do not modify my main conclusions.


judge, who is still in charge of the pre-trial investigation;288 and at trial, thecourt decides what evidence will be produced and in what order,289 and hasan active role in the interrogation of witnesses, including expert witnesses.290

Therefore, despite the procedimiento abreviado’s similarity to U.S. sentencingbargaining, the model of the ofªcial investigation is still much more pre-dominant in Argentina than in Italy. While the dispute model has appearedin Argentine federal criminal procedures, its presence is still minor.

From the perspective of the Americanization thesis debate, Argentine fed-eral criminal procedure has not been Americanized; the model of the ofªcialinvestigation is still overwhelmingly the conception of criminal procedurewithin Argentine legal practices. The main reason for this seems to be thateven if the translation of U.S. plea bargaining has remained quite faithful tothe original text, very few other reforms with a potential adversarializationeffect have been introduced. The differences between the adversarial and theinquisitorial cultures are so deep that the Americanization effect of a singleadversarial reform like the procedimiento abreviado are likely to be easily neu-tralized. The wooden horse has been brought within the walls of the Argen-tine criminal procedure, but the soldiers in its belly have not been enough totake Troy, or even a substantial part of the city. Only time will tell whethernew Trojan horses will appear in the future.

X. French “Plea Bargaining”291

French “plea bargaining” or composition hardly resembles the Americanmodel. It differs from U.S. practice in that it is more of a tool for divertingcases from formal criminal proceedings than a way of disposing of cases afterthey have entered such proceedings. Therefore, this plea bargaining transla-tion arguably will not move the predominantly inquisitorial French criminalprocedure in the direction of the American adversarial system.

288. See Cód. Proc. Pen. art. 26 (Arg.).289. See Cód. Proc. Pen. art. 356 (Arg.).290. See Cód. Proc. Pen. art. 389 (Arg.).291. The French Minister of Justice, Dominique Perben, has presented a bill that is pending before

the French Congress (as of Oct. 24, 2003), proposing substantial reforms to French criminal procedure.See Projet de loi portant adaptation de la justice aux évolutions de la criminalité at http://www.assemblee-nat.fr/12/pdf/projets/pl0784.pdf (last visited Dec. 3, 2003). This proposed statute, known as the “Perben Bill”(“projet de loi Perben”), has been presented as an attempt at “adaptation of the administration of justice tothe evolution of criminality.” This proposal includes the possibility of the defendant pleading guilty forcrimes of up to ªve years of imprisonment. Since this is just a proposal that has not yet been approved, Ido not analyze it in detail in this Article. Another attempt to introduce a mechanism similar to guiltypleas happened at the end of the 1980s and the beginning of the 1990s, when a Commission of theFrench Minister of Justice—usually known as the Delmas-Marty Commission—proposed the introduc-tion of a mechanism similar to the U.S. guilty plea, though it speciªcally rejected the introduction ofplea bargaining. See Commission Justice Pénale et Droits de l’Homme, La “Mise en État” des

Affaires Pénales (1991). Two of the reasons that French legal actors rejected these proposals were thatthey saw them as foreign and that they had an attachment to the existing inquisitorial system (see DenisSalas, La justice entre les deux “corps” de la démocratie, in La justice, une révolution démocratique 20(Denis Salas ed., 2001). For a brief analysis of the proposal of the Delmas-Marty Commission regardingguilty pleas, see Tulkens, supra note 178, at 672.


At the end of June 1999, the composition was introduced in arts. 41-2 and41-3 of the French Criminal Procedure Code,292 with the goal of reducingthe caseload of the courts.293 This mechanism has been considered a homo-logue of U.S. plea bargaining because it gives room for negotiations betweenthe prosecution and the defense.294 According to the composition, before thebeginning of the formal proceedings, the prosecution may offer the defen-dant the option of diverting his case from the standard criminal trial in ex-change for an admission of guilt and the fulªllment of a condition such aspaying a ªne, turning over any objects used to commit the offense (or ob-jects obtained in the course of the offense), forfeiting his driving or huntinglicense for a certain period of time, doing community service work, and/orrepairing the damage done to the victim.295 If the defendant accepts the of-fer, the prosecutor requests that it be validated by the judge.296 If the defen-dant does not accept the offer, or does not fulªll the conditions of theagreement, the prosecutor can simply initiate the formal proceedings.297

The composition can only be applied to certain offenses speciªcally listed inthe French Code, such as simple assault, threats, simple robbery, criminaldamages, criminal libel and slander, cruelty against animals, possession ofcertain weapons, or driving while intoxicated, among others.298 Further, it isto be applied only to non-serious offenses.299

292. Law No. 99-515 of June 23, 1999, J.O., June 24, 1999, p. 9207. The predecessor of the

composition pénale was the injonction pénale voted by law on December 22, 1994. However, on February 2,1995, the Constitutional Court (Conseil Constitutionnel) held the injonction unconstitutional saying thatit went against the presumption of innocence and that it gave the prosecutor the power to try cases andimpose sanctions that should be imposed by the courts. See Merle & Vitu, supra note 109, at 396.

293. See Philippe Conte & Patrick Maistre du Chambon, Procédure Pénale, 3–4 (3d ed.2001); Merle & Vitu, supra note 109, at 395.

294. Merle & Vitu, supra note 109, at 396; Cedras, supra note 179, at 156–57; Jean-ClaudeMagendie, L’américanisation du droit ou la création d’un mythe, 45 Archives de Philosophie du Droit

13, 18–19 (R. Sève et al. eds., 2001). There are other mechanisms in French criminal procedure that giveroom for negotiations between the prosecutor, the defendant, and even the victim and the judge, such asthe médiation pénale, correctionnalisation, and comparution immédiate. See Tulkens, supra note 178, at 660–61,672–73. I will not analyze them here, and, to my knowledge, they have not been inspired by, orconsidered analogous to, plea bargaining. For analyses of the role of negotiation not only in criminal lawbut also in other legal ªelds, see, generally, the articles included in Droit Négocié, Droit Imposé

(Philippe Gérard et al. eds., 1996).295. C. pr. pén. art. 41–42 (2002) (France). For a detailed analysis of this mechanism, see Ministère

de la Justice, Circulaires de la direction des Affaires criminelles et des Grâces, 83 Bulletin Ofªciel du

Ministère de la Justice 25 (2001), available at http://www.justice.gouv.fr/actua/bo/dacg83c.htm (lastvisited Nov. 25, 2003). The proposal cannot be made when the defendant is under arrest. The prosecutoror his representative has to inform the defendant of his right to be assisted by an attorney.

296. The defendant and the victim alike can request a hearing from the judge before she decidesabout the agreement. However, this hearing is not conducted if it is not speciªcally requested. See C. pr.

pén. art. 41–42 (Fr.). Because the composition was introduced as a way to deal more effectively with minorcrime, it is assumed that a hearing with the defendant and/or the victim before validating the agreementis exceptional. Ministère de la Justice, supra note 295, at 96.

297. While the conditions of the agreement are being fulªlled, the statute of limitations is suspended.298. C. pr. pén. art. 41–42 (Fr.); Ministère de la Justice, supra note 295, at 69–70.299. Ministère de la Justice, supra note 295, at 68. Besides proposing the introduction of a new

mechanism similar to the American guilty plea (reconnaissance préalable de culpabilité), see Perben, infra note300, the “Perben Bill” also proposes an extension of the composition to all offenses of up to ªve years of


As is already clear from this description, one could only refer to the com-position as “plea bargaining” in a very loose way.300 Here, again, it is clearhow inadequate the idea of the legal transplant is to capture the phenome-non of circulating legal ideas and practices. There are some similarities be-tween the procedures of American plea bargaining and the composition: bothcan include negotiations between the prosecutor and the defendant, and thelatter has to admit his guilt as part of the agreement. This, however, is theextent of their similarity.

The differences between U.S. plea bargaining and the composition aresigniªcant. First, while the application of plea bargaining shortens the regularcriminal proceedings in the sense that a trial is not necessary to determineguilt or innocence, the application of the composition directly avoids such pro-ceedings.301 Second, though the application of American plea bargainingmeans that the defendant is found legally guilty and then punished, the ap-plication of the composition does not have the legal effect of a guilty verdict. Ifthe defendant fulªlls the conditions of the agreement, the case is dismissedwith prejudice.302 Third, in plea bargaining, the prosecutor is understood tobe in an equal bargaining position with the defense. In the composition, theprosecutor does not negotiate with an equal but is more akin to a diversionofªcer to exert control over a person who has broken the law and may com-mit new offenses in the future. The defendant must accept the prosecutor’soffer and admit his guilt, not as a party who can end the dispute with hisconsent, but rather as part of his own process of neutralization, rehabilita-tion, and reparation to the victim.303

Through the composition, French legislators have translated U.S. plea bar-gaining not from the adversarial procedural structure of meaning to an in-quisitorial one, but rather from the adversarial system to what can be calledthe “depenalization model.” In the second half of the twentieth century, sev-eral European countries witnessed a movement toward the depenalization of

imprisonment. See Dominique Perben, Projet de loi portant adaptation de la justice aux évolutions de la crimi-nalité, tit. II, ch. 1, sec. 2 (“Dispositions relatives à la composition pénale et aux autres procédures alter-natives aux poursuites”) (proposing a new text for art. 41-2.1 of the French Code of Criminal Procedure),available at http://www.assemblee-nat.fr/12/projets/pl0784.asp (last visited Nov. 25, 2003) (on ªle withthe Harvard International Law Journal).

300. The mechanism proposed by the “Perben Bill,” see supra note 295, is much more similar to theAmerican plea bargain because it includes negotiations between the prosecution and the defense regard-ing punishment in exchange for an admission of guilt within the formal criminal proceedings, and thecourt has a limited role in the negotiations. See Dominique Perben, Projet de loi portant adaptation de lajustice aux évolutions de la criminalité, ch. IV, sec. 1, art. 61, sec. VIII (“De la comparution sur reconnais-sance préalable de culpabilité”), available at http://www.assemblee-nat.fr/12/projets/pl0784.asp (lastvisited Nov. 25, 2003) (on ªle with the Harvard International Law Journal). If ultimately adopted, the“Perben Bill” could potentially have an “adversarialization effect” that the composition does not seem tohave.

301. See Merle & Vitu, supra note 109, at 396.302. C. pr. pén. art. 41–42 (2002) (Fr.).303. The admission of guilt by the defendant has a pedagogic character. Ministère de la Justice,

supra note 295, at 65.


less serious offenses.304 According to the depenalization model, the regularcriminal justice system, with its imprisonment penalties, is too harsh and atthe same time ineffective in dealing with non-serious criminality. Fines,community work, and reparation to the victim, among other remedies, havebeen proposed to replace imprisonment. A number of diversion systems havealso been designed to allow administrative, civil, and mediation courts tohandle cases that were traditionally adjudicated through the criminal proc-ess.305

The depenalization model does not attempt to abolish or entirely replacethe criminal justice system and its proceedings. Rather, it tries to limit theirscope while functioning in a complementary fashion.306 It does not adoptany speciªc kind of procedure. It may work comfortably with relatively in-quisitorial administrative proceedings, with civil proceedings structuredaccording to the adversarial model of the dispute, or with mediation pro-grams. The depenalization model aims to take less serious offenses out offormal criminal proceedings and to avoid recourse to imprisonment, while atthe same time maintaining a certain formal control over these cases; there-fore, there is a multiplicity of procedures through which these aims can beachieved.

The composition can be included within this depenalization model.307

Therefore, even if the composition is ªnally successful and widely used,308 it isunlikely that it will move the regular French inquisitorial criminal pro-ceedings in the direction of the model of the dispute; rather, it is likely tostrengthen the depenalization model within France. First, as we alreadyanalyzed, the composition itself, even if inspired by plea bargaining, does notassume the model of the dispute, mainly because it does not treat the defen-dant as a party equal to the prosecution. In this sense, the composition doesnot appear to have a potential adversarialization or Americanization effect.Second, even if it did, the very fact that the composition is not applied within

304. For an analysis of the development of this model in Europe, see Mireille Delmas-Marty, Les

grands systèmes de politique criminelle 159–63, 278–93 (1992). In the United States, there havebeen certain tendencies in the same direction. For instance, the introduction of diversion mechanisms inrecent years for dealing with drug, domestic violence, and juvenile cases could be considered part of asimilar development. For an analysis of the expansion of diversion in the United States in the last threedecades, see, for example, Miller et al., supra note 76, at 663–65.

305. See, e.g., Tulkens, supra note 178, at 648–49.306. Professor Delmas-Marty makes a distinction between depenalization and decriminalization.

While the former could serve as a recourse complementary to the criminal justice system, the latterwould present itself as an alternative to it. See Delmas-Marty, supra note 304, at 278–79.

307. The composition is part of the depenalization model because it attempts to take certain cases out ofthe regular criminal justice system while keeping these cases under a soft criminal justice control. Onthis double goal of the composition, see Conte & Maistre du Chambon, supra note 293, at 209.

308. So far, empirical evidence shows that the composition has been used sparingly. There are ofªcialstatistics about the use of this practice only for the year 2001. Out of 893,373 cases prosecuted, thecomposition was used in only 1511 (0.17%). See Ministère de la Justice, Annuaire Statistique de la

Justice 2003, available at http://www.justice.gouv.fr/publicat/m-Activparquets.pdf (last visited Nov. 25,2003) (on ªle with the Harvard International Law Journal).


regular criminal proceedings would also make its potential to move theFrench model of the ofªcial investigation in that direction unlikely.

The translation of plea bargaining to France, then, does not seem to repre-sent a challenge to its predominant model of the ofªcial investigation incriminal procedure, even less of a challenge than the introduction of the pro-cedimiento abreviado to Argentina.

XI. Conclusion: Fragmentation in the Civil Law?

In the last four Parts, I have shown how, in recent times, the civil law ju-risdictions of Germany, Italy, Argentina, and France have translated Ameri-can plea bargaining into their jurisdictions in different ways and how thesemechanisms, in their interactions with the legal actors and practices of thereceiving criminal justice system and other factors, may be producing differ-ent outcomes in each of these jurisdictions.

In the case of Germany, the Absprachen brought the development of a newcriminal procedural system—the model of the judge-manager based on co-operation—that now co-exists and challenges the traditional German in-quisitorial model of the ofªcial investigation. In the case of Italy, the patteg-giamento, acting jointly with other reforms, introduced the model of the dis-pute that now represents a serious challenge to the traditionally predomi-nant model of the ofªcial investigation. In the case of Argentina, the procedi-miento abreviado has introduced and reinforced elements of the model of thedispute in federal criminal procedure, though the model of the ofªcial inves-tigation is still predominant, much more so than in Italy. Finally, in the caseof France, the composition may reinforce the depenalization model and reducethe range of cases processed through the model of the ofªcial investigation.Nevertheless, this last model is still highly predominant in regular criminalproceedings.

This analysis demonstrates that there now appear to be important differ-ences among these jurisdictions and that these differences are somethingnew within a civil law world that has long been a relatively homogeneouslegal culture. While there were differences among civil law jurisdictions,and challenges to the inquisitorial procedural culture within each of them,in all four jurisdictions examined here, criminal procedure was still over-whelmingly conceptualized around the model of the ofªcial investigation. Itis on this deep common conception of criminal procedure that these fourcivil law jurisdictions are beginning to differ.

Thus, while American inºuences on the civil law world have been unde-niable, at least in its formal criminal procedures, they are not producing astrong Americanization, or adversarialization, of the civil law, but rather itsfragmentation. This fragmentation is due at least in part to the fact thatinquisitorial systems have “translated” American adversarial inºuences indifferent ways. In its interactions with the receiving criminal procedure,each translated mechanism has the potential to both transform and be


transformed by the inquisitorial system in different ways. As a result, thecriminal procedures of the civil law tradition have already begun a process ofdifferentiation.

This phenomenon has important consequences in the debate on the Ameri-canization thesis. While it supports the weak version of the Americanizationthesis in that it substantiates the adversarial inºuences on inquisitorial sys-tems, it should also lead to a revision and a more concrete and subtle analy-sis of the strong version of this thesis. As shown by this discussion, theinºuences of the American system on other legal systems may produce dif-ferent effects in each of them depending on the decisions made by legaltranslators, how much power they have had to advance their reforms, howmuch resistance there has been against them, etc. Furthermore, this studydemonstrates that Americanization is not an all-or-nothing game. In otherwords, the practices of the receiving legal system may move in the directionof the American system at one level of procedure, but not at another, as inthe case of the German system, where the Absprachen is unlikely to advancethe American model of the dispute but may advance a more consensual andrelative conception of truth like that which prevails in the American system.

In making my argument, I have also made two additional points worthreiterating here. First, I have discussed some of the shortcomings of themetaphor of the legal transplant as a means of understanding the circulationof legal institutions between legal systems, and I have proposed the meta-phor of the legal translation as an alternative heuristic device to approachthese issues. Some may consider this discussion of metaphors as a word gamewithout much substantive value, but this would be a mistake. Metaphorsmatter because we think through them and because they highlight differentaspects of reality. The main concept that I seek to emphasize through themetaphor of the legal translation is that legal practices and institutions maybe transformed when translated between legal systems either because of de-cisions by the reformers (translators) or structural differences between theoriginal and receiving legal systems (languages). Furthermore, the translatedlegal institution may produce a chain of changes—even unexpected ones asin the case of the German Absprachen—in the receiving legal system. Giventhat these transformations have been overlooked by many scholars and pol-icy-makers, the translation metaphor is useful to highlight such changes.

Second, I have shown that the adversarial and inquisitorial systems can beanalyzed not only as two different ways to distribute powers and responsi-bilities between the main actors of the criminal process, but also as two pro-cedural cultures—two different conceptions of how criminal cases should betried and prosecuted. Even if these procedural cultures are not entirely ho-mogeneous, certain conceptions of criminal procedure have clearly prevailed inthe United States and in the four civil law jurisdictions examined here. Ihave proposed a new theoretical framework to analyze these predominantprocedural cultures and have used it to explain some of the transformationsthat plea bargaining has undergone when translated to the civil law jurisdic-


tions of Germany, Italy, Argentina, and France. This conceptualization of theadversarial and the inquisitorial systems has also provided us with a clearaxis of reference to assess whether our four civil law jurisdictions are movingor have moved in the direction of the American system.

The conceptualization of the adversarial and the inquisitorial systems asprocedural cultures is also important for other reasons. The debate aboutAmericanization of law is, to a great extent, a debate about legal cultures. Inother words, it is a debate about how law is understood, thought of, andpracticed in different jurisdictions, as well as about how certain conceptionsof legal phenomena that prevail in the United States may overcome otherconceptions. A theoretical conceptualization of the adversarial and inquisito-rial systems of the sort offered in this Article is thus necessary to fully par-ticipate in any such debate.

from legal transplants to legal translations: the ... · americanization thesis in criminal...

Documents